01a_machine_learning.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Creative machine learning - Machine Learning\n",
    "\n",
    "### Author: Philippe Esling (esling@ircam.fr)\n",
    "\n",
    "In this course we will cover\n",
    "1. A [first definition](#definition) on the concept of machine learning\n",
    "2. An introduction to a simple problem of [linear regression](#regression)\n",
    "4. A detailed implementation of [simple linear regression](#linear)\n",
    "3. An explanation on [model capacity and overfitting](#capacity)\n",
    "4. Two [exercises](#exercises) to explore regression and classification\n",
    "4. An introduction to the [audio datasets](#audio) that we will use"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"definition\"></a>\n",
    "## Defining machine learning\n",
    "\n",
    "In all natural process, there exists complex relations between sets $\\mathcal{X} \\mapsto \\mathcal{Y}$. This can relate some objects with their names, or a cause to a consequence. In most cases, _we do not know the precise relations_ between these sets, all we have is _observations_ such as pairs $(x,y)$, composed of input data $x \\in \\mathcal{X}$, which have a corresponding expected output $y \\in \\mathcal{Y}$. The overarching goal of machine learning is to approximate such _unknown processes_ as a function $\\mathcal{F}_{\\theta}$, which _transforms_ input data $x$ into output data $y$.\n",
    "\n",
    "\n",
    "<center>\n",
    "<img src=\"images/01_machine_learning_basic.png\" align=\"center\"/>\n",
    "</center>\n",
    "\n",
    "Hence, machine learning aims to understand and model the relationship between some (usually complex and high-dimensional) inputs $\\mathbf{x}\\in\\mathcal{X}\\subset\\mathbb{R}^{\\mathcal{X}}$ and outputs $\\mathbf{y}\\in\\mathcal{Y}\\subset\\mathbb{R}^{\\mathcal{Y}}$, given by a set of data examples $\\mathcal{D}=\\left\\{(x_1,y_1),\\cdots,(x_N,y_N)\\right\\}$. This is achieved by defining a parametric model $f_{\\mathbf{\\theta}}\\in\\mathcal{F}$ inside a family of functions $\\mathcal{F}$, which depends on parameters $\\mathbf{\\theta} \\in \\mathbf{\\Theta}$ and that could approximate the underlying relationship. The _learning_ aspect refers to the adjustment of the parameters $\\mathbf{\\theta}$ in order to obtain the best approximation of the given task\n",
    "$$\n",
    "\\begin{equation}\n",
    "f_{\\mathbf{\\theta}}(\\mathbf{x}) = \\bar{\\mathbf{y}}\\approx \\mathbf{y}.\n",
    "\\end{equation}\n",
    "$$\n",
    "\n",
    "Hence, the major elements that we have to define in any machine learning problems are\n",
    "1. **Dataset** : $\\mathcal{D}=\\left\\{(x_1,y_1),\\cdots,(x_N,y_N)\\right\\}$. This dataset has to be representative of the relation $f:\\mathcal{X} \\mapsto \\mathcal{Y}$ that we are looking to model\n",
    "2. **Model** : Our parametric approximation $\\bar{\\mathbf{y}} = f_{\\mathbf{\\theta}}(\\mathbf{x})$, where the choice of family $f_{\\mathbf{\\theta}}\\in\\mathcal{F}$ is critical\n",
    "3. **Loss** : $\\mathcal{L}\\left( \\bar{\\mathbf{y}}, \\mathbf{y} \\mid f_{\\theta}, \\theta \\right)$ allows to measure the amount of errors made by our model\n",
    "4. **Optimization** : Method to find $\\theta^{*}\\in\\Theta$ so that our model minimizes the loss\n",
    "$$\\theta^{*}= \\underset{\\theta}{\\text{argmin }} \\mathcal{L}\\left( \\bar{\\mathbf{y}}, \\mathbf{y} \\mid f_{\\theta}, \\theta \\right)$$\n",
    "\n",
    "\n",
    "To observe this idea in simple setups, we are going to use the `numpy` library and also initialize the homemade course library `cml` and style for future plotting and exercise. We also set the random generator to a fixed point with `rng = np.random.RandomState(1)`, to ensure reproducibility"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div class=\"bk-root\">\n",
       "        <a href=\"https://bokeh.org\" target=\"_blank\" class=\"bk-logo bk-logo-small bk-logo-notebook\"></a>\n",
       "        <span id=\"1002\">Loading BokehJS ...</span>\n",
       "    </div>\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "(function(root) {\n",
       "  function now() {\n",
       "    return new Date();\n",
       "  }\n",
       "\n",
       "  const force = true;\n",
       "\n",
       "  if (typeof root._bokeh_onload_callbacks === \"undefined\" || force === true) {\n",
       "    root._bokeh_onload_callbacks = [];\n",
       "    root._bokeh_is_loading = undefined;\n",
       "  }\n",
       "\n",
       "const JS_MIME_TYPE = 'application/javascript';\n",
       "  const HTML_MIME_TYPE = 'text/html';\n",
       "  const EXEC_MIME_TYPE = 'application/vnd.bokehjs_exec.v0+json';\n",
       "  const CLASS_NAME = 'output_bokeh rendered_html';\n",
       "\n",
       "  /**\n",
       "   * Render data to the DOM node\n",
       "   */\n",
       "  function render(props, node) {\n",
       "    const script = document.createElement(\"script\");\n",
       "    node.appendChild(script);\n",
       "  }\n",
       "\n",
       "  /**\n",
       "   * Handle when an output is cleared or removed\n",
       "   */\n",
       "  function handleClearOutput(event, handle) {\n",
       "    const cell = handle.cell;\n",
       "\n",
       "    const id = cell.output_area._bokeh_element_id;\n",
       "    const server_id = cell.output_area._bokeh_server_id;\n",
       "    // Clean up Bokeh references\n",
       "    if (id != null && id in Bokeh.index) {\n",
       "      Bokeh.index[id].model.document.clear();\n",
       "      delete Bokeh.index[id];\n",
       "    }\n",
       "\n",
       "    if (server_id !== undefined) {\n",
       "      // Clean up Bokeh references\n",
       "      const cmd_clean = \"from bokeh.io.state import curstate; print(curstate().uuid_to_server['\" + server_id + \"'].get_sessions()[0].document.roots[0]._id)\";\n",
       "      cell.notebook.kernel.execute(cmd_clean, {\n",
       "        iopub: {\n",
       "          output: function(msg) {\n",
       "            const id = msg.content.text.trim();\n",
       "            if (id in Bokeh.index) {\n",
       "              Bokeh.index[id].model.document.clear();\n",
       "              delete Bokeh.index[id];\n",
       "            }\n",
       "          }\n",
       "        }\n",
       "      });\n",
       "      // Destroy server and session\n",
       "      const cmd_destroy = \"import bokeh.io.notebook as ion; ion.destroy_server('\" + server_id + \"')\";\n",
       "      cell.notebook.kernel.execute(cmd_destroy);\n",
       "    }\n",
       "  }\n",
       "\n",
       "  /**\n",
       "   * Handle when a new output is added\n",
       "   */\n",
       "  function handleAddOutput(event, handle) {\n",
       "    const output_area = handle.output_area;\n",
       "    const output = handle.output;\n",
       "\n",
       "    // limit handleAddOutput to display_data with EXEC_MIME_TYPE content only\n",
       "    if ((output.output_type != \"display_data\") || (!Object.prototype.hasOwnProperty.call(output.data, EXEC_MIME_TYPE))) {\n",
       "      return\n",
       "    }\n",
       "\n",
       "    const toinsert = output_area.element.find(\".\" + CLASS_NAME.split(' ')[0]);\n",
       "\n",
       "    if (output.metadata[EXEC_MIME_TYPE][\"id\"] !== undefined) {\n",
       "      toinsert[toinsert.length - 1].firstChild.textContent = output.data[JS_MIME_TYPE];\n",
       "      // store reference to embed id on output_area\n",
       "      output_area._bokeh_element_id = output.metadata[EXEC_MIME_TYPE][\"id\"];\n",
       "    }\n",
       "    if (output.metadata[EXEC_MIME_TYPE][\"server_id\"] !== undefined) {\n",
       "      const bk_div = document.createElement(\"div\");\n",
       "      bk_div.innerHTML = output.data[HTML_MIME_TYPE];\n",
       "      const script_attrs = bk_div.children[0].attributes;\n",
       "      for (let i = 0; i < script_attrs.length; i++) {\n",
       "        toinsert[toinsert.length - 1].firstChild.setAttribute(script_attrs[i].name, script_attrs[i].value);\n",
       "        toinsert[toinsert.length - 1].firstChild.textContent = bk_div.children[0].textContent\n",
       "      }\n",
       "      // store reference to server id on output_area\n",
       "      output_area._bokeh_server_id = output.metadata[EXEC_MIME_TYPE][\"server_id\"];\n",
       "    }\n",
       "  }\n",
       "\n",
       "  function register_renderer(events, OutputArea) {\n",
       "\n",
       "    function append_mime(data, metadata, element) {\n",
       "      // create a DOM node to render to\n",
       "      const toinsert = this.create_output_subarea(\n",
       "        metadata,\n",
       "        CLASS_NAME,\n",
       "        EXEC_MIME_TYPE\n",
       "      );\n",
       "      this.keyboard_manager.register_events(toinsert);\n",
       "      // Render to node\n",
       "      const props = {data: data, metadata: metadata[EXEC_MIME_TYPE]};\n",
       "      render(props, toinsert[toinsert.length - 1]);\n",
       "      element.append(toinsert);\n",
       "      return toinsert\n",
       "    }\n",
       "\n",
       "    /* Handle when an output is cleared or removed */\n",
       "    events.on('clear_output.CodeCell', handleClearOutput);\n",
       "    events.on('delete.Cell', handleClearOutput);\n",
       "\n",
       "    /* Handle when a new output is added */\n",
       "    events.on('output_added.OutputArea', handleAddOutput);\n",
       "\n",
       "    /**\n",
       "     * Register the mime type and append_mime function with output_area\n",
       "     */\n",
       "    OutputArea.prototype.register_mime_type(EXEC_MIME_TYPE, append_mime, {\n",
       "      /* Is output safe? */\n",
       "      safe: true,\n",
       "      /* Index of renderer in `output_area.display_order` */\n",
       "      index: 0\n",
       "    });\n",
       "  }\n",
       "\n",
       "  // register the mime type if in Jupyter Notebook environment and previously unregistered\n",
       "  if (root.Jupyter !== undefined) {\n",
       "    const events = require('base/js/events');\n",
       "    const OutputArea = require('notebook/js/outputarea').OutputArea;\n",
       "\n",
       "    if (OutputArea.prototype.mime_types().indexOf(EXEC_MIME_TYPE) == -1) {\n",
       "      register_renderer(events, OutputArea);\n",
       "    }\n",
       "  }\n",
       "  if (typeof (root._bokeh_timeout) === \"undefined\" || force === true) {\n",
       "    root._bokeh_timeout = Date.now() + 5000;\n",
       "    root._bokeh_failed_load = false;\n",
       "  }\n",
       "\n",
       "  const NB_LOAD_WARNING = {'data': {'text/html':\n",
       "     \"<div style='background-color: #fdd'>\\n\"+\n",
       "     \"<p>\\n\"+\n",
       "     \"BokehJS does not appear to have successfully loaded. If loading BokehJS from CDN, this \\n\"+\n",
       "     \"may be due to a slow or bad network connection. Possible fixes:\\n\"+\n",
       "     \"</p>\\n\"+\n",
       "     \"<ul>\\n\"+\n",
       "     \"<li>re-rerun `output_notebook()` to attempt to load from CDN again, or</li>\\n\"+\n",
       "     \"<li>use INLINE resources instead, as so:</li>\\n\"+\n",
       "     \"</ul>\\n\"+\n",
       "     \"<code>\\n\"+\n",
       "     \"from bokeh.resources import INLINE\\n\"+\n",
       "     \"output_notebook(resources=INLINE)\\n\"+\n",
       "     \"</code>\\n\"+\n",
       "     \"</div>\"}};\n",
       "\n",
       "  function display_loaded() {\n",
       "    const el = document.getElementById(\"1002\");\n",
       "    if (el != null) {\n",
       "      el.textContent = \"BokehJS is loading...\";\n",
       "    }\n",
       "    if (root.Bokeh !== undefined) {\n",
       "      if (el != null) {\n",
       "        el.textContent = \"BokehJS \" + root.Bokeh.version + \" successfully loaded.\";\n",
       "      }\n",
       "    } else if (Date.now() < root._bokeh_timeout) {\n",
       "      setTimeout(display_loaded, 100)\n",
       "    }\n",
       "  }\n",
       "\n",
       "  function run_callbacks() {\n",
       "    try {\n",
       "      root._bokeh_onload_callbacks.forEach(function(callback) {\n",
       "        if (callback != null)\n",
       "          callback();\n",
       "      });\n",
       "    } finally {\n",
       "      delete root._bokeh_onload_callbacks\n",
       "    }\n",
       "    console.debug(\"Bokeh: all callbacks have finished\");\n",
       "  }\n",
       "\n",
       "  function load_libs(css_urls, js_urls, callback) {\n",
       "    if (css_urls == null) css_urls = [];\n",
       "    if (js_urls == null) js_urls = [];\n",
       "\n",
       "    root._bokeh_onload_callbacks.push(callback);\n",
       "    if (root._bokeh_is_loading > 0) {\n",
       "      console.debug(\"Bokeh: BokehJS is being loaded, scheduling callback at\", now());\n",
       "      return null;\n",
       "    }\n",
       "    if (js_urls == null || js_urls.length === 0) {\n",
       "      run_callbacks();\n",
       "      return null;\n",
       "    }\n",
       "    console.debug(\"Bokeh: BokehJS not loaded, scheduling load and callback at\", now());\n",
       "    root._bokeh_is_loading = css_urls.length + js_urls.length;\n",
       "\n",
       "    function on_load() {\n",
       "      root._bokeh_is_loading--;\n",
       "      if (root._bokeh_is_loading === 0) {\n",
       "        console.debug(\"Bokeh: all BokehJS libraries/stylesheets loaded\");\n",
       "        run_callbacks()\n",
       "      }\n",
       "    }\n",
       "\n",
       "    function on_error(url) {\n",
       "      console.error(\"failed to load \" + url);\n",
       "    }\n",
       "\n",
       "    for (let i = 0; i < css_urls.length; i++) {\n",
       "      const url = css_urls[i];\n",
       "      const element = document.createElement(\"link\");\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error.bind(null, url);\n",
       "      element.rel = \"stylesheet\";\n",
       "      element.type = \"text/css\";\n",
       "      element.href = url;\n",
       "      console.debug(\"Bokeh: injecting link tag for BokehJS stylesheet: \", url);\n",
       "      document.body.appendChild(element);\n",
       "    }\n",
       "\n",
       "    for (let i = 0; i < js_urls.length; i++) {\n",
       "      const url = js_urls[i];\n",
       "      const element = document.createElement('script');\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error.bind(null, url);\n",
       "      element.async = false;\n",
       "      element.src = url;\n",
       "      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      document.head.appendChild(element);\n",
       "    }\n",
       "  };\n",
       "\n",
       "  function inject_raw_css(css) {\n",
       "    const element = document.createElement(\"style\");\n",
       "    element.appendChild(document.createTextNode(css));\n",
       "    document.body.appendChild(element);\n",
       "  }\n",
       "\n",
       "  const js_urls = [\"https://cdn.bokeh.org/bokeh/release/bokeh-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-gl-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-widgets-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-tables-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-mathjax-2.4.3.min.js\", \"https://unpkg.com/@holoviz/panel@0.14.4/dist/panel.min.js\"];\n",
       "  const css_urls = [];\n",
       "\n",
       "  const inline_js = [    function(Bokeh) {\n",
       "      Bokeh.set_log_level(\"info\");\n",
       "    },\n",
       "function(Bokeh) {\n",
       "    }\n",
       "  ];\n",
       "\n",
       "  function run_inline_js() {\n",
       "    if (root.Bokeh !== undefined || force === true) {\n",
       "          for (let i = 0; i < inline_js.length; i++) {\n",
       "      inline_js[i].call(root, root.Bokeh);\n",
       "    }\n",
       "if (force === true) {\n",
       "        display_loaded();\n",
       "      }} else if (Date.now() < root._bokeh_timeout) {\n",
       "      setTimeout(run_inline_js, 100);\n",
       "    } else if (!root._bokeh_failed_load) {\n",
       "      console.log(\"Bokeh: BokehJS failed to load within specified timeout.\");\n",
       "      root._bokeh_failed_load = true;\n",
       "    } else if (force !== true) {\n",
       "      const cell = $(document.getElementById(\"1002\")).parents('.cell').data().cell;\n",
       "      cell.output_area.append_execute_result(NB_LOAD_WARNING)\n",
       "    }\n",
       "  }\n",
       "\n",
       "  if (root._bokeh_is_loading === 0) {\n",
       "    console.debug(\"Bokeh: BokehJS loaded, going straight to plotting\");\n",
       "    run_inline_js();\n",
       "  } else {\n",
       "    load_libs(css_urls, js_urls, function() {\n",
       "      console.debug(\"Bokeh: BokehJS plotting callback run at\", now());\n",
       "      run_inline_js();\n",
       "    });\n",
       "  }\n",
       "}(window));"
      ],
      "application/vnd.bokehjs_load.v0+json": "(function(root) {\n  function now() {\n    return new Date();\n  }\n\n  const force = true;\n\n  if (typeof root._bokeh_onload_callbacks === \"undefined\" || force === true) {\n    root._bokeh_onload_callbacks = [];\n    root._bokeh_is_loading = undefined;\n  }\n\n\n  if (typeof (root._bokeh_timeout) === \"undefined\" || force === true) {\n    root._bokeh_timeout = Date.now() + 5000;\n    root._bokeh_failed_load = false;\n  }\n\n  const NB_LOAD_WARNING = {'data': {'text/html':\n     \"<div style='background-color: #fdd'>\\n\"+\n     \"<p>\\n\"+\n     \"BokehJS does not appear to have successfully loaded. If loading BokehJS from CDN, this \\n\"+\n     \"may be due to a slow or bad network connection. Possible fixes:\\n\"+\n     \"</p>\\n\"+\n     \"<ul>\\n\"+\n     \"<li>re-rerun `output_notebook()` to attempt to load from CDN again, or</li>\\n\"+\n     \"<li>use INLINE resources instead, as so:</li>\\n\"+\n     \"</ul>\\n\"+\n     \"<code>\\n\"+\n     \"from bokeh.resources import INLINE\\n\"+\n     \"output_notebook(resources=INLINE)\\n\"+\n     \"</code>\\n\"+\n     \"</div>\"}};\n\n  function display_loaded() {\n    const el = document.getElementById(\"1002\");\n    if (el != null) {\n      el.textContent = \"BokehJS is loading...\";\n    }\n    if (root.Bokeh !== undefined) {\n      if (el != null) {\n        el.textContent = \"BokehJS \" + root.Bokeh.version + \" successfully loaded.\";\n      }\n    } else if (Date.now() < root._bokeh_timeout) {\n      setTimeout(display_loaded, 100)\n    }\n  }\n\n  function run_callbacks() {\n    try {\n      root._bokeh_onload_callbacks.forEach(function(callback) {\n        if (callback != null)\n          callback();\n      });\n    } finally {\n      delete root._bokeh_onload_callbacks\n    }\n    console.debug(\"Bokeh: all callbacks have finished\");\n  }\n\n  function load_libs(css_urls, js_urls, callback) {\n    if (css_urls == null) css_urls = [];\n    if (js_urls == null) js_urls = [];\n\n    root._bokeh_onload_callbacks.push(callback);\n    if (root._bokeh_is_loading > 0) {\n      console.debug(\"Bokeh: BokehJS is being loaded, scheduling callback at\", now());\n      return null;\n    }\n    if (js_urls == null || js_urls.length === 0) {\n      run_callbacks();\n      return null;\n    }\n    console.debug(\"Bokeh: BokehJS not loaded, scheduling load and callback at\", now());\n    root._bokeh_is_loading = css_urls.length + js_urls.length;\n\n    function on_load() {\n      root._bokeh_is_loading--;\n      if (root._bokeh_is_loading === 0) {\n        console.debug(\"Bokeh: all BokehJS libraries/stylesheets loaded\");\n        run_callbacks()\n      }\n    }\n\n    function on_error(url) {\n      console.error(\"failed to load \" + url);\n    }\n\n    for (let i = 0; i < css_urls.length; i++) {\n      const url = css_urls[i];\n      const element = document.createElement(\"link\");\n      element.onload = on_load;\n      element.onerror = on_error.bind(null, url);\n      element.rel = \"stylesheet\";\n      element.type = \"text/css\";\n      element.href = url;\n      console.debug(\"Bokeh: injecting link tag for BokehJS stylesheet: \", url);\n      document.body.appendChild(element);\n    }\n\n    for (let i = 0; i < js_urls.length; i++) {\n      const url = js_urls[i];\n      const element = document.createElement('script');\n      element.onload = on_load;\n      element.onerror = on_error.bind(null, url);\n      element.async = false;\n      element.src = url;\n      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n      document.head.appendChild(element);\n    }\n  };\n\n  function inject_raw_css(css) {\n    const element = document.createElement(\"style\");\n    element.appendChild(document.createTextNode(css));\n    document.body.appendChild(element);\n  }\n\n  const js_urls = [\"https://cdn.bokeh.org/bokeh/release/bokeh-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-gl-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-widgets-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-tables-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-mathjax-2.4.3.min.js\", \"https://unpkg.com/@holoviz/panel@0.14.4/dist/panel.min.js\"];\n  const css_urls = [];\n\n  const inline_js = [    function(Bokeh) {\n      Bokeh.set_log_level(\"info\");\n    },\nfunction(Bokeh) {\n    }\n  ];\n\n  function run_inline_js() {\n    if (root.Bokeh !== undefined || force === true) {\n          for (let i = 0; i < inline_js.length; i++) {\n      inline_js[i].call(root, root.Bokeh);\n    }\nif (force === true) {\n        display_loaded();\n      }} else if (Date.now() < root._bokeh_timeout) {\n      setTimeout(run_inline_js, 100);\n    } else if (!root._bokeh_failed_load) {\n      console.log(\"Bokeh: BokehJS failed to load within specified timeout.\");\n      root._bokeh_failed_load = true;\n    } else if (force !== true) {\n      const cell = $(document.getElementById(\"1002\")).parents('.cell').data().cell;\n      cell.output_area.append_execute_result(NB_LOAD_WARNING)\n    }\n  }\n\n  if (root._bokeh_is_loading === 0) {\n    console.debug(\"Bokeh: BokehJS loaded, going straight to plotting\");\n    run_inline_js();\n  } else {\n    load_libs(css_urls, js_urls, function() {\n      console.debug(\"Bokeh: BokehJS plotting callback run at\", now());\n      run_inline_js();\n    });\n  }\n}(window));"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "(function(root) {\n",
       "  function now() {\n",
       "    return new Date();\n",
       "  }\n",
       "\n",
       "  var force = true;\n",
       "\n",
       "  if (typeof root._bokeh_onload_callbacks === \"undefined\" || force === true) {\n",
       "    root._bokeh_onload_callbacks = [];\n",
       "    root._bokeh_is_loading = undefined;\n",
       "  }\n",
       "\n",
       "  if (typeof (root._bokeh_timeout) === \"undefined\" || force === true) {\n",
       "    root._bokeh_timeout = Date.now() + 5000;\n",
       "    root._bokeh_failed_load = false;\n",
       "  }\n",
       "\n",
       "  function run_callbacks() {\n",
       "    try {\n",
       "      root._bokeh_onload_callbacks.forEach(function(callback) {\n",
       "        if (callback != null)\n",
       "          callback();\n",
       "      });\n",
       "    } finally {\n",
       "      delete root._bokeh_onload_callbacks\n",
       "    }\n",
       "    console.debug(\"Bokeh: all callbacks have finished\");\n",
       "  }\n",
       "\n",
       "  function load_libs(css_urls, js_urls, js_modules, callback) {\n",
       "    if (css_urls == null) css_urls = [];\n",
       "    if (js_urls == null) js_urls = [];\n",
       "    if (js_modules == null) js_modules = [];\n",
       "\n",
       "    root._bokeh_onload_callbacks.push(callback);\n",
       "    if (root._bokeh_is_loading > 0) {\n",
       "      console.debug(\"Bokeh: BokehJS is being loaded, scheduling callback at\", now());\n",
       "      return null;\n",
       "    }\n",
       "    if (js_urls.length === 0 && js_modules.length === 0) {\n",
       "      run_callbacks();\n",
       "      return null;\n",
       "    }\n",
       "    console.debug(\"Bokeh: BokehJS not loaded, scheduling load and callback at\", now());\n",
       "\n",
       "    function on_load() {\n",
       "      root._bokeh_is_loading--;\n",
       "      if (root._bokeh_is_loading === 0) {\n",
       "        console.debug(\"Bokeh: all BokehJS libraries/stylesheets loaded\");\n",
       "        run_callbacks()\n",
       "      }\n",
       "    }\n",
       "\n",
       "    function on_error() {\n",
       "      console.error(\"failed to load \" + url);\n",
       "    }\n",
       "\n",
       "    for (var i = 0; i < css_urls.length; i++) {\n",
       "      var url = css_urls[i];\n",
       "      const element = document.createElement(\"link\");\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error;\n",
       "      element.rel = \"stylesheet\";\n",
       "      element.type = \"text/css\";\n",
       "      element.href = url;\n",
       "      console.debug(\"Bokeh: injecting link tag for BokehJS stylesheet: \", url);\n",
       "      document.body.appendChild(element);\n",
       "    }\n",
       "\n",
       "    var skip = [];\n",
       "    if (window.requirejs) {\n",
       "      window.requirejs.config({'packages': {}, 'paths': {'ace': '//cdnjs.cloudflare.com/ajax/libs/ace/1.4.7', 'gridstack': 'https://cdn.jsdelivr.net/npm/gridstack@4.2.5/dist/gridstack-h5', 'notyf': 'https://cdn.jsdelivr.net/npm/notyf@3/notyf.min'}, 'shim': {'ace/ext-language_tools': {'deps': ['ace/ace']}, 'ace/ext-modelist': {'deps': ['ace/ace']}, 'gridstack': {'exports': 'GridStack'}}});\n",
       "      require([\"ace/ace\"], function(ace) {\n",
       "\twindow.ace = ace\n",
       "\ton_load()\n",
       "      })\n",
       "      require([\"ace/ext-language_tools\"], function() {\n",
       "\ton_load()\n",
       "      })\n",
       "      require([\"ace/ext-modelist\"], function() {\n",
       "\ton_load()\n",
       "      })\n",
       "      require([\"gridstack\"], function(GridStack) {\n",
       "\twindow.GridStack = GridStack\n",
       "\ton_load()\n",
       "      })\n",
       "      require([\"notyf\"], function() {\n",
       "\ton_load()\n",
       "      })\n",
       "      root._bokeh_is_loading = css_urls.length + 5;\n",
       "    } else {\n",
       "      root._bokeh_is_loading = css_urls.length + js_urls.length + js_modules.length;\n",
       "    }    if (((window['ace'] !== undefined) && (!(window['ace'] instanceof HTMLElement))) || window.requirejs) {\n",
       "      var urls = ['https://cdnjs.cloudflare.com/ajax/libs/ace/1.4.11/ace.js', 'https://cdnjs.cloudflare.com/ajax/libs/ace/1.4.11/ext-language_tools.js', 'https://cdnjs.cloudflare.com/ajax/libs/ace/1.4.11/ext-modelist.js'];\n",
       "      for (var i = 0; i < urls.length; i++) {\n",
       "        skip.push(urls[i])\n",
       "      }\n",
       "    }    if (((window['GridStack'] !== undefined) && (!(window['GridStack'] instanceof HTMLElement))) || window.requirejs) {\n",
       "      var urls = ['https://cdn.holoviz.org/panel/0.14.4/dist/bundled/gridstack/gridstack@4.2.5/dist/gridstack-h5.js'];\n",
       "      for (var i = 0; i < urls.length; i++) {\n",
       "        skip.push(urls[i])\n",
       "      }\n",
       "    }    if (((window['Notyf'] !== undefined) && (!(window['Notyf'] instanceof HTMLElement))) || window.requirejs) {\n",
       "      var urls = ['https://cdn.holoviz.org/panel/0.14.4/dist/bundled/notificationarea/notyf@3/notyf.min.js'];\n",
       "      for (var i = 0; i < urls.length; i++) {\n",
       "        skip.push(urls[i])\n",
       "      }\n",
       "    }    for (var i = 0; i < js_urls.length; i++) {\n",
       "      var url = js_urls[i];\n",
       "      if (skip.indexOf(url) >= 0) {\n",
       "\tif (!window.requirejs) {\n",
       "\t  on_load();\n",
       "\t}\n",
       "\tcontinue;\n",
       "      }\n",
       "      var element = document.createElement('script');\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error;\n",
       "      element.async = false;\n",
       "      element.src = url;\n",
       "      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      document.head.appendChild(element);\n",
       "    }\n",
       "    for (var i = 0; i < js_modules.length; i++) {\n",
       "      var url = js_modules[i];\n",
       "      if (skip.indexOf(url) >= 0) {\n",
       "\tif (!window.requirejs) {\n",
       "\t  on_load();\n",
       "\t}\n",
       "\tcontinue;\n",
       "      }\n",
       "      var element = document.createElement('script');\n",
       "      element.onload = on_load;\n",
       "      element.onerror = on_error;\n",
       "      element.async = false;\n",
       "      element.src = url;\n",
       "      element.type = \"module\";\n",
       "      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      document.head.appendChild(element);\n",
       "    }\n",
       "    if (!js_urls.length && !js_modules.length) {\n",
       "      on_load()\n",
       "    }\n",
       "  };\n",
       "\n",
       "  function inject_raw_css(css) {\n",
       "    const element = document.createElement(\"style\");\n",
       "    element.appendChild(document.createTextNode(css));\n",
       "    document.body.appendChild(element);\n",
       "  }\n",
       "\n",
       "  var js_urls = [\"https://cdnjs.cloudflare.com/ajax/libs/ace/1.4.11/ace.js\", \"https://cdnjs.cloudflare.com/ajax/libs/ace/1.4.11/ext-language_tools.js\", \"https://cdnjs.cloudflare.com/ajax/libs/ace/1.4.11/ext-modelist.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-gl-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-widgets-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-tables-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-mathjax-2.4.3.min.js\", \"https://unpkg.com/@holoviz/panel@0.14.4/dist/panel.min.js\"];\n",
       "  var js_modules = [];\n",
       "  var css_urls = [\"https://cdn.holoviz.org/panel/0.14.4/dist/css/debugger.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/widgets.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/json.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/dataframe.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/card.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/loading.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/markdown.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/alerts.css\"];\n",
       "  var inline_js = [    function(Bokeh) {\n",
       "      inject_raw_css(\"/*\\n ~ CML // Creative Machine Learning ~\\n mml.css : CSS styling information for Panel and Bokeh\\n \\n This file defines the main CSS styling information for the CML course\\n \\n Author               :  Philippe Esling\\n                        <esling@ircam.fr>\\n*/\\n\\nbody {\\n  display: flex;\\n  height: 100vh;\\n  margin: 0px;\\n  overflow-x: hidden;\\n  overflow-y: hidden;\\n}\\n\\n.bk-root .bk, .bk-root .bk:before, .bk-root .bk:after {\\n  font-family: \\\"Josefin Sans\\\";\\n}\\n\\nimg {\\n  max-width: 100%;\\n}\\n\\n#container {\\n  padding:0px;\\n  height:100vh;\\n  width: 100vw;\\n  max-width: 100vw;\\n}\\n\\n#sidebar .mdc-list {\\n  padding-left: 5px;\\n  padding-right: 5px;\\n}\\n\\n.mdc-drawer-app-content {\\n  flex: auto;\\n  position: relative;\\n  overflow: hidden;\\n}\\n\\n.mdc-drawer {\\n  background: #FAFAFA; /* GRAY 50 */\\n}\\n\\n.mdc-drawer-app-content {\\n  margin-left: 0 !important;\\n}\\n\\n.title-bar {\\n  display: contents;\\n  justify-content: center;\\n  align-content: center;\\n  width: 100%;\\n}\\n\\n.mdc-top-app-bar .bk-menu {\\n  color: black\\n}\\n\\n.app-header {\\n  display: contents;\\n  padding-left: 10px;\\n  font-size: 1.25em;\\n}\\n\\nimg.app-logo {\\n  padding-right: 10px;\\n  font-size: 28px;\\n  height: 30px;\\n  max-width: inherit;\\n  padding-top: 12px;\\n  padding-bottom: 6px;\\n}\\n\\n#app-title {\\n  padding-right: 12px;\\n  padding-left: 12px;\\n}\\n\\n.title {\\n  font-family: \\\"Josefin Sans\\\";\\n  color: #fff;\\n  text-decoration: none;\\n  text-decoration-line: none;\\n  text-decoration-style: initial;\\n  text-decoration-color: initial;\\n  font-weight: 400;\\n  font-size: 2em;\\n  line-height: 2em;\\n  white-space: nowrap;\\n}\\n\\n.main-content {\\n  overflow-y: scroll;\\n  overflow-x: auto;\\n}\\n\\n#header {\\n  position: absolute;\\n  z-index: 7;\\n}\\n\\n#header-items {\\n  width: 100%;\\n  margin-left:15px;\\n}\\n\\n.pn-busy-container {\\n  align-items: center;\\n  justify-content: center;\\n  display: flex;\\n}\\n\\n.mdc-drawer__content {\\n  overflow-x: hidden;\\n}\\n.mdc-drawer__content, .main-content {\\n  padding: 12px;\\n}\\n\\n.main-content {\\n  height: calc(100vh - 88px);\\n  max-height: calc(100vh - 88px);\\n  padding-right: 32px;\\n}\\n\\nbutton.mdc-button.mdc-card-button {\\n  color: transparent;\\n  height: 50px;\\n}\\n\\np.bk.mdc-button {\\n  display: none;\\n}\\n\\ndiv.bk.mdc-card {\\n  border-radius: 0px\\n}\\n\\n.mdc-card .bk.card-header {\\n  display: flex;\\n}\\n\\n.bk.mdc-card-title {\\n  font-family: \\\"Josefin Sans\\\";\\n  font-weight: bold;\\n  align-items: center;\\n  display: flex !important;\\n  position: relative !important;\\n}\\n\\n.bk.mdc-card-title:nth-child(2) {\\n  margin-left: -1.4em;\\n}\\n\\n.pn-modal {\\n  overflow-y: scroll;\\n  width: 100%;\\n  display: none;\\n  position: absolute;\\n  top: 0;\\n  left: 0;\\n}\\n\\n.pn-modal-content {\\n  font-family: \\\"Josefin Sans\\\";\\n  background-color: #0e0e0e;\\n  margin: auto;\\n  margin-top: 25px;\\n  margin-bottom: 25px;\\n  padding: 15px 20px 20px 20px;\\n  border: 1px solid #888;\\n  width: 80% !important;\\n}\\n\\n.pn-modal-close {\\n  position: absolute;\\n  right: 25px;\\n  z-index: 100;\\n}\\n\\n.pn-modal-close:hover,\\n.pn-modal-close:focus {\\n  color: #000;\\n  text-decoration: none;\\n  cursor: pointer;\\n}\\n\\n.custom_button_bokeh button.bk.bk-btn.bk-btn-default {\\n    font-size:48pt;\\n    background-color: #05b7ff;\\n    border-color: #05b7ff;\\n}\");\n",
       "    },    function(Bokeh) {\n",
       "      inject_raw_css(\"\\n    .bk.pn-loading.arc:before {\\n      background-image: url(\\\"data:image/svg+xml;base64,PHN2ZyB4bWxucz0iaHR0cDovL3d3dy53My5vcmcvMjAwMC9zdmciIHN0eWxlPSJtYXJnaW46IGF1dG87IGJhY2tncm91bmQ6IG5vbmU7IGRpc3BsYXk6IGJsb2NrOyBzaGFwZS1yZW5kZXJpbmc6IGF1dG87IiB2aWV3Qm94PSIwIDAgMTAwIDEwMCIgcHJlc2VydmVBc3BlY3RSYXRpbz0ieE1pZFlNaWQiPiAgPGNpcmNsZSBjeD0iNTAiIGN5PSI1MCIgZmlsbD0ibm9uZSIgc3Ryb2tlPSIjYzNjM2MzIiBzdHJva2Utd2lkdGg9IjEwIiByPSIzNSIgc3Ryb2tlLWRhc2hhcnJheT0iMTY0LjkzMzYxNDMxMzQ2NDE1IDU2Ljk3Nzg3MTQzNzgyMTM4Ij4gICAgPGFuaW1hdGVUcmFuc2Zvcm0gYXR0cmlidXRlTmFtZT0idHJhbnNmb3JtIiB0eXBlPSJyb3RhdGUiIHJlcGVhdENvdW50PSJpbmRlZmluaXRlIiBkdXI9IjFzIiB2YWx1ZXM9IjAgNTAgNTA7MzYwIDUwIDUwIiBrZXlUaW1lcz0iMDsxIj48L2FuaW1hdGVUcmFuc2Zvcm0+ICA8L2NpcmNsZT48L3N2Zz4=\\\");\\n      background-size: auto calc(min(50%, 400px));\\n    }\\n    \");\n",
       "    },    function(Bokeh) {\n",
       "      Bokeh.set_log_level(\"info\");\n",
       "    },\n",
       "function(Bokeh) {} // ensure no trailing comma for IE\n",
       "  ];\n",
       "\n",
       "  function run_inline_js() {\n",
       "    if ((root.Bokeh !== undefined) || (force === true)) {\n",
       "      for (var i = 0; i < inline_js.length; i++) {\n",
       "        inline_js[i].call(root, root.Bokeh);\n",
       "      }} else if (Date.now() < root._bokeh_timeout) {\n",
       "      setTimeout(run_inline_js, 100);\n",
       "    } else if (!root._bokeh_failed_load) {\n",
       "      console.log(\"Bokeh: BokehJS failed to load within specified timeout.\");\n",
       "      root._bokeh_failed_load = true;\n",
       "    }\n",
       "  }\n",
       "\n",
       "  if (root._bokeh_is_loading === 0) {\n",
       "    console.debug(\"Bokeh: BokehJS loaded, going straight to plotting\");\n",
       "    run_inline_js();\n",
       "  } else {\n",
       "    load_libs(css_urls, js_urls, js_modules, function() {\n",
       "      console.debug(\"Bokeh: BokehJS plotting callback run at\", now());\n",
       "      run_inline_js();\n",
       "    });\n",
       "  }\n",
       "}(window));"
      ],
      "application/vnd.holoviews_load.v0+json": "(function(root) {\n  function now() {\n    return new Date();\n  }\n\n  var force = true;\n\n  if (typeof root._bokeh_onload_callbacks === \"undefined\" || force === true) {\n    root._bokeh_onload_callbacks = [];\n    root._bokeh_is_loading = undefined;\n  }\n\n  if (typeof (root._bokeh_timeout) === \"undefined\" || force === true) {\n    root._bokeh_timeout = Date.now() + 5000;\n    root._bokeh_failed_load = false;\n  }\n\n  function run_callbacks() {\n    try {\n      root._bokeh_onload_callbacks.forEach(function(callback) {\n        if (callback != null)\n          callback();\n      });\n    } finally {\n      delete root._bokeh_onload_callbacks\n    }\n    console.debug(\"Bokeh: all callbacks have finished\");\n  }\n\n  function load_libs(css_urls, js_urls, js_modules, callback) {\n    if (css_urls == null) css_urls = [];\n    if (js_urls == null) js_urls = [];\n    if (js_modules == null) js_modules = [];\n\n    root._bokeh_onload_callbacks.push(callback);\n    if (root._bokeh_is_loading > 0) {\n      console.debug(\"Bokeh: BokehJS is being loaded, scheduling callback at\", now());\n      return null;\n    }\n    if (js_urls.length === 0 && js_modules.length === 0) {\n      run_callbacks();\n      return null;\n    }\n    console.debug(\"Bokeh: BokehJS not loaded, scheduling load and callback at\", now());\n\n    function on_load() {\n      root._bokeh_is_loading--;\n      if (root._bokeh_is_loading === 0) {\n        console.debug(\"Bokeh: all BokehJS libraries/stylesheets loaded\");\n        run_callbacks()\n      }\n    }\n\n    function on_error() {\n      console.error(\"failed to load \" + url);\n    }\n\n    for (var i = 0; i < css_urls.length; i++) {\n      var url = css_urls[i];\n      const element = document.createElement(\"link\");\n      element.onload = on_load;\n      element.onerror = on_error;\n      element.rel = \"stylesheet\";\n      element.type = \"text/css\";\n      element.href = url;\n      console.debug(\"Bokeh: injecting link tag for BokehJS stylesheet: \", url);\n      document.body.appendChild(element);\n    }\n\n    var skip = [];\n    if (window.requirejs) {\n      window.requirejs.config({'packages': {}, 'paths': {'ace': '//cdnjs.cloudflare.com/ajax/libs/ace/1.4.7', 'gridstack': 'https://cdn.jsdelivr.net/npm/gridstack@4.2.5/dist/gridstack-h5', 'notyf': 'https://cdn.jsdelivr.net/npm/notyf@3/notyf.min'}, 'shim': {'ace/ext-language_tools': {'deps': ['ace/ace']}, 'ace/ext-modelist': {'deps': ['ace/ace']}, 'gridstack': {'exports': 'GridStack'}}});\n      require([\"ace/ace\"], function(ace) {\n\twindow.ace = ace\n\ton_load()\n      })\n      require([\"ace/ext-language_tools\"], function() {\n\ton_load()\n      })\n      require([\"ace/ext-modelist\"], function() {\n\ton_load()\n      })\n      require([\"gridstack\"], function(GridStack) {\n\twindow.GridStack = GridStack\n\ton_load()\n      })\n      require([\"notyf\"], function() {\n\ton_load()\n      })\n      root._bokeh_is_loading = css_urls.length + 5;\n    } else {\n      root._bokeh_is_loading = css_urls.length + js_urls.length + js_modules.length;\n    }    if (((window['ace'] !== undefined) && (!(window['ace'] instanceof HTMLElement))) || window.requirejs) {\n      var urls = ['https://cdnjs.cloudflare.com/ajax/libs/ace/1.4.11/ace.js', 'https://cdnjs.cloudflare.com/ajax/libs/ace/1.4.11/ext-language_tools.js', 'https://cdnjs.cloudflare.com/ajax/libs/ace/1.4.11/ext-modelist.js'];\n      for (var i = 0; i < urls.length; i++) {\n        skip.push(urls[i])\n      }\n    }    if (((window['GridStack'] !== undefined) && (!(window['GridStack'] instanceof HTMLElement))) || window.requirejs) {\n      var urls = ['https://cdn.holoviz.org/panel/0.14.4/dist/bundled/gridstack/gridstack@4.2.5/dist/gridstack-h5.js'];\n      for (var i = 0; i < urls.length; i++) {\n        skip.push(urls[i])\n      }\n    }    if (((window['Notyf'] !== undefined) && (!(window['Notyf'] instanceof HTMLElement))) || window.requirejs) {\n      var urls = ['https://cdn.holoviz.org/panel/0.14.4/dist/bundled/notificationarea/notyf@3/notyf.min.js'];\n      for (var i = 0; i < urls.length; i++) {\n        skip.push(urls[i])\n      }\n    }    for (var i = 0; i < js_urls.length; i++) {\n      var url = js_urls[i];\n      if (skip.indexOf(url) >= 0) {\n\tif (!window.requirejs) {\n\t  on_load();\n\t}\n\tcontinue;\n      }\n      var element = document.createElement('script');\n      element.onload = on_load;\n      element.onerror = on_error;\n      element.async = false;\n      element.src = url;\n      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n      document.head.appendChild(element);\n    }\n    for (var i = 0; i < js_modules.length; i++) {\n      var url = js_modules[i];\n      if (skip.indexOf(url) >= 0) {\n\tif (!window.requirejs) {\n\t  on_load();\n\t}\n\tcontinue;\n      }\n      var element = document.createElement('script');\n      element.onload = on_load;\n      element.onerror = on_error;\n      element.async = false;\n      element.src = url;\n      element.type = \"module\";\n      console.debug(\"Bokeh: injecting script tag for BokehJS library: \", url);\n      document.head.appendChild(element);\n    }\n    if (!js_urls.length && !js_modules.length) {\n      on_load()\n    }\n  };\n\n  function inject_raw_css(css) {\n    const element = document.createElement(\"style\");\n    element.appendChild(document.createTextNode(css));\n    document.body.appendChild(element);\n  }\n\n  var js_urls = [\"https://cdnjs.cloudflare.com/ajax/libs/ace/1.4.11/ace.js\", \"https://cdnjs.cloudflare.com/ajax/libs/ace/1.4.11/ext-language_tools.js\", \"https://cdnjs.cloudflare.com/ajax/libs/ace/1.4.11/ext-modelist.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-gl-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-widgets-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-tables-2.4.3.min.js\", \"https://cdn.bokeh.org/bokeh/release/bokeh-mathjax-2.4.3.min.js\", \"https://unpkg.com/@holoviz/panel@0.14.4/dist/panel.min.js\"];\n  var js_modules = [];\n  var css_urls = [\"https://cdn.holoviz.org/panel/0.14.4/dist/css/debugger.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/widgets.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/json.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/dataframe.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/card.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/loading.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/markdown.css\", \"https://cdn.holoviz.org/panel/0.14.4/dist/css/alerts.css\"];\n  var inline_js = [    function(Bokeh) {\n      inject_raw_css(\"/*\\n ~ CML // Creative Machine Learning ~\\n mml.css : CSS styling information for Panel and Bokeh\\n \\n This file defines the main CSS styling information for the CML course\\n \\n Author               :  Philippe Esling\\n                        <esling@ircam.fr>\\n*/\\n\\nbody {\\n  display: flex;\\n  height: 100vh;\\n  margin: 0px;\\n  overflow-x: hidden;\\n  overflow-y: hidden;\\n}\\n\\n.bk-root .bk, .bk-root .bk:before, .bk-root .bk:after {\\n  font-family: \\\"Josefin Sans\\\";\\n}\\n\\nimg {\\n  max-width: 100%;\\n}\\n\\n#container {\\n  padding:0px;\\n  height:100vh;\\n  width: 100vw;\\n  max-width: 100vw;\\n}\\n\\n#sidebar .mdc-list {\\n  padding-left: 5px;\\n  padding-right: 5px;\\n}\\n\\n.mdc-drawer-app-content {\\n  flex: auto;\\n  position: relative;\\n  overflow: hidden;\\n}\\n\\n.mdc-drawer {\\n  background: #FAFAFA; /* GRAY 50 */\\n}\\n\\n.mdc-drawer-app-content {\\n  margin-left: 0 !important;\\n}\\n\\n.title-bar {\\n  display: contents;\\n  justify-content: center;\\n  align-content: center;\\n  width: 100%;\\n}\\n\\n.mdc-top-app-bar .bk-menu {\\n  color: black\\n}\\n\\n.app-header {\\n  display: contents;\\n  padding-left: 10px;\\n  font-size: 1.25em;\\n}\\n\\nimg.app-logo {\\n  padding-right: 10px;\\n  font-size: 28px;\\n  height: 30px;\\n  max-width: inherit;\\n  padding-top: 12px;\\n  padding-bottom: 6px;\\n}\\n\\n#app-title {\\n  padding-right: 12px;\\n  padding-left: 12px;\\n}\\n\\n.title {\\n  font-family: \\\"Josefin Sans\\\";\\n  color: #fff;\\n  text-decoration: none;\\n  text-decoration-line: none;\\n  text-decoration-style: initial;\\n  text-decoration-color: initial;\\n  font-weight: 400;\\n  font-size: 2em;\\n  line-height: 2em;\\n  white-space: nowrap;\\n}\\n\\n.main-content {\\n  overflow-y: scroll;\\n  overflow-x: auto;\\n}\\n\\n#header {\\n  position: absolute;\\n  z-index: 7;\\n}\\n\\n#header-items {\\n  width: 100%;\\n  margin-left:15px;\\n}\\n\\n.pn-busy-container {\\n  align-items: center;\\n  justify-content: center;\\n  display: flex;\\n}\\n\\n.mdc-drawer__content {\\n  overflow-x: hidden;\\n}\\n.mdc-drawer__content, .main-content {\\n  padding: 12px;\\n}\\n\\n.main-content {\\n  height: calc(100vh - 88px);\\n  max-height: calc(100vh - 88px);\\n  padding-right: 32px;\\n}\\n\\nbutton.mdc-button.mdc-card-button {\\n  color: transparent;\\n  height: 50px;\\n}\\n\\np.bk.mdc-button {\\n  display: none;\\n}\\n\\ndiv.bk.mdc-card {\\n  border-radius: 0px\\n}\\n\\n.mdc-card .bk.card-header {\\n  display: flex;\\n}\\n\\n.bk.mdc-card-title {\\n  font-family: \\\"Josefin Sans\\\";\\n  font-weight: bold;\\n  align-items: center;\\n  display: flex !important;\\n  position: relative !important;\\n}\\n\\n.bk.mdc-card-title:nth-child(2) {\\n  margin-left: -1.4em;\\n}\\n\\n.pn-modal {\\n  overflow-y: scroll;\\n  width: 100%;\\n  display: none;\\n  position: absolute;\\n  top: 0;\\n  left: 0;\\n}\\n\\n.pn-modal-content {\\n  font-family: \\\"Josefin Sans\\\";\\n  background-color: #0e0e0e;\\n  margin: auto;\\n  margin-top: 25px;\\n  margin-bottom: 25px;\\n  padding: 15px 20px 20px 20px;\\n  border: 1px solid #888;\\n  width: 80% !important;\\n}\\n\\n.pn-modal-close {\\n  position: absolute;\\n  right: 25px;\\n  z-index: 100;\\n}\\n\\n.pn-modal-close:hover,\\n.pn-modal-close:focus {\\n  color: #000;\\n  text-decoration: none;\\n  cursor: pointer;\\n}\\n\\n.custom_button_bokeh button.bk.bk-btn.bk-btn-default {\\n    font-size:48pt;\\n    background-color: #05b7ff;\\n    border-color: #05b7ff;\\n}\");\n    },    function(Bokeh) {\n      inject_raw_css(\"\\n    .bk.pn-loading.arc:before {\\n      background-image: url(\\\"data:image/svg+xml;base64,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\\\");\\n      background-size: auto calc(min(50%, 400px));\\n    }\\n    \");\n    },    function(Bokeh) {\n      Bokeh.set_log_level(\"info\");\n    },\nfunction(Bokeh) {} // ensure no trailing comma for IE\n  ];\n\n  function run_inline_js() {\n    if ((root.Bokeh !== undefined) || (force === true)) {\n      for (var i = 0; i < inline_js.length; i++) {\n        inline_js[i].call(root, root.Bokeh);\n      }} else if (Date.now() < root._bokeh_timeout) {\n      setTimeout(run_inline_js, 100);\n    } else if (!root._bokeh_failed_load) {\n      console.log(\"Bokeh: BokehJS failed to load within specified timeout.\");\n      root._bokeh_failed_load = true;\n    }\n  }\n\n  if (root._bokeh_is_loading === 0) {\n    console.debug(\"Bokeh: BokehJS loaded, going straight to plotting\");\n    run_inline_js();\n  } else {\n    load_libs(css_urls, js_urls, js_modules, function() {\n      console.debug(\"Bokeh: BokehJS plotting callback run at\", now());\n      run_inline_js();\n    });\n  }\n}(window));"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "if ((window.PyViz === undefined) || (window.PyViz instanceof HTMLElement)) {\n",
       "  window.PyViz = {comms: {}, comm_status:{}, kernels:{}, receivers: {}, plot_index: []}\n",
       "}\n",
       "\n",
       "\n",
       "    function JupyterCommManager() {\n",
       "    }\n",
       "\n",
       "    JupyterCommManager.prototype.register_target = function(plot_id, comm_id, msg_handler) {\n",
       "      if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n",
       "        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n",
       "        comm_manager.register_target(comm_id, function(comm) {\n",
       "          comm.on_msg(msg_handler);\n",
       "        });\n",
       "      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n",
       "        window.PyViz.kernels[plot_id].registerCommTarget(comm_id, function(comm) {\n",
       "          comm.onMsg = msg_handler;\n",
       "        });\n",
       "      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n",
       "        google.colab.kernel.comms.registerTarget(comm_id, (comm) => {\n",
       "          var messages = comm.messages[Symbol.asyncIterator]();\n",
       "          function processIteratorResult(result) {\n",
       "            var message = result.value;\n",
       "            console.log(message)\n",
       "            var content = {data: message.data, comm_id};\n",
       "            var buffers = []\n",
       "            for (var buffer of message.buffers || []) {\n",
       "              buffers.push(new DataView(buffer))\n",
       "            }\n",
       "            var metadata = message.metadata || {};\n",
       "            var msg = {content, buffers, metadata}\n",
       "            msg_handler(msg);\n",
       "            return messages.next().then(processIteratorResult);\n",
       "          }\n",
       "          return messages.next().then(processIteratorResult);\n",
       "        })\n",
       "      }\n",
       "    }\n",
       "\n",
       "    JupyterCommManager.prototype.get_client_comm = function(plot_id, comm_id, msg_handler) {\n",
       "      if (comm_id in window.PyViz.comms) {\n",
       "        return window.PyViz.comms[comm_id];\n",
       "      } else if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n",
       "        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n",
       "        var comm = comm_manager.new_comm(comm_id, {}, {}, {}, comm_id);\n",
       "        if (msg_handler) {\n",
       "          comm.on_msg(msg_handler);\n",
       "        }\n",
       "      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n",
       "        var comm = window.PyViz.kernels[plot_id].connectToComm(comm_id);\n",
       "        comm.open();\n",
       "        if (msg_handler) {\n",
       "          comm.onMsg = msg_handler;\n",
       "        }\n",
       "      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n",
       "        var comm_promise = google.colab.kernel.comms.open(comm_id)\n",
       "        comm_promise.then((comm) => {\n",
       "          window.PyViz.comms[comm_id] = comm;\n",
       "          if (msg_handler) {\n",
       "            var messages = comm.messages[Symbol.asyncIterator]();\n",
       "            function processIteratorResult(result) {\n",
       "              var message = result.value;\n",
       "              var content = {data: message.data};\n",
       "              var metadata = message.metadata || {comm_id};\n",
       "              var msg = {content, metadata}\n",
       "              msg_handler(msg);\n",
       "              return messages.next().then(processIteratorResult);\n",
       "            }\n",
       "            return messages.next().then(processIteratorResult);\n",
       "          }\n",
       "        }) \n",
       "        var sendClosure = (data, metadata, buffers, disposeOnDone) => {\n",
       "          return comm_promise.then((comm) => {\n",
       "            comm.send(data, metadata, buffers, disposeOnDone);\n",
       "          });\n",
       "        };\n",
       "        var comm = {\n",
       "          send: sendClosure\n",
       "        };\n",
       "      }\n",
       "      window.PyViz.comms[comm_id] = comm;\n",
       "      return comm;\n",
       "    }\n",
       "    window.PyViz.comm_manager = new JupyterCommManager();\n",
       "    \n",
       "\n",
       "\n",
       "var JS_MIME_TYPE = 'application/javascript';\n",
       "var HTML_MIME_TYPE = 'text/html';\n",
       "var EXEC_MIME_TYPE = 'application/vnd.holoviews_exec.v0+json';\n",
       "var CLASS_NAME = 'output';\n",
       "\n",
       "/**\n",
       " * Render data to the DOM node\n",
       " */\n",
       "function render(props, node) {\n",
       "  var div = document.createElement(\"div\");\n",
       "  var script = document.createElement(\"script\");\n",
       "  node.appendChild(div);\n",
       "  node.appendChild(script);\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle when a new output is added\n",
       " */\n",
       "function handle_add_output(event, handle) {\n",
       "  var output_area = handle.output_area;\n",
       "  var output = handle.output;\n",
       "  if ((output.data == undefined) || (!output.data.hasOwnProperty(EXEC_MIME_TYPE))) {\n",
       "    return\n",
       "  }\n",
       "  var id = output.metadata[EXEC_MIME_TYPE][\"id\"];\n",
       "  var toinsert = output_area.element.find(\".\" + CLASS_NAME.split(' ')[0]);\n",
       "  if (id !== undefined) {\n",
       "    var nchildren = toinsert.length;\n",
       "    var html_node = toinsert[nchildren-1].children[0];\n",
       "    html_node.innerHTML = output.data[HTML_MIME_TYPE];\n",
       "    var scripts = [];\n",
       "    var nodelist = html_node.querySelectorAll(\"script\");\n",
       "    for (var i in nodelist) {\n",
       "      if (nodelist.hasOwnProperty(i)) {\n",
       "        scripts.push(nodelist[i])\n",
       "      }\n",
       "    }\n",
       "\n",
       "    scripts.forEach( function (oldScript) {\n",
       "      var newScript = document.createElement(\"script\");\n",
       "      var attrs = [];\n",
       "      var nodemap = oldScript.attributes;\n",
       "      for (var j in nodemap) {\n",
       "        if (nodemap.hasOwnProperty(j)) {\n",
       "          attrs.push(nodemap[j])\n",
       "        }\n",
       "      }\n",
       "      attrs.forEach(function(attr) { newScript.setAttribute(attr.name, attr.value) });\n",
       "      newScript.appendChild(document.createTextNode(oldScript.innerHTML));\n",
       "      oldScript.parentNode.replaceChild(newScript, oldScript);\n",
       "    });\n",
       "    if (JS_MIME_TYPE in output.data) {\n",
       "      toinsert[nchildren-1].children[1].textContent = output.data[JS_MIME_TYPE];\n",
       "    }\n",
       "    output_area._hv_plot_id = id;\n",
       "    if ((window.Bokeh !== undefined) && (id in Bokeh.index)) {\n",
       "      window.PyViz.plot_index[id] = Bokeh.index[id];\n",
       "    } else {\n",
       "      window.PyViz.plot_index[id] = null;\n",
       "    }\n",
       "  } else if (output.metadata[EXEC_MIME_TYPE][\"server_id\"] !== undefined) {\n",
       "    var bk_div = document.createElement(\"div\");\n",
       "    bk_div.innerHTML = output.data[HTML_MIME_TYPE];\n",
       "    var script_attrs = bk_div.children[0].attributes;\n",
       "    for (var i = 0; i < script_attrs.length; i++) {\n",
       "      toinsert[toinsert.length - 1].childNodes[1].setAttribute(script_attrs[i].name, script_attrs[i].value);\n",
       "    }\n",
       "    // store reference to server id on output_area\n",
       "    output_area._bokeh_server_id = output.metadata[EXEC_MIME_TYPE][\"server_id\"];\n",
       "  }\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle when an output is cleared or removed\n",
       " */\n",
       "function handle_clear_output(event, handle) {\n",
       "  var id = handle.cell.output_area._hv_plot_id;\n",
       "  var server_id = handle.cell.output_area._bokeh_server_id;\n",
       "  if (((id === undefined) || !(id in PyViz.plot_index)) && (server_id !== undefined)) { return; }\n",
       "  var comm = window.PyViz.comm_manager.get_client_comm(\"hv-extension-comm\", \"hv-extension-comm\", function () {});\n",
       "  if (server_id !== null) {\n",
       "    comm.send({event_type: 'server_delete', 'id': server_id});\n",
       "    return;\n",
       "  } else if (comm !== null) {\n",
       "    comm.send({event_type: 'delete', 'id': id});\n",
       "  }\n",
       "  delete PyViz.plot_index[id];\n",
       "  if ((window.Bokeh !== undefined) & (id in window.Bokeh.index)) {\n",
       "    var doc = window.Bokeh.index[id].model.document\n",
       "    doc.clear();\n",
       "    const i = window.Bokeh.documents.indexOf(doc);\n",
       "    if (i > -1) {\n",
       "      window.Bokeh.documents.splice(i, 1);\n",
       "    }\n",
       "  }\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle kernel restart event\n",
       " */\n",
       "function handle_kernel_cleanup(event, handle) {\n",
       "  delete PyViz.comms[\"hv-extension-comm\"];\n",
       "  window.PyViz.plot_index = {}\n",
       "}\n",
       "\n",
       "/**\n",
       " * Handle update_display_data messages\n",
       " */\n",
       "function handle_update_output(event, handle) {\n",
       "  handle_clear_output(event, {cell: {output_area: handle.output_area}})\n",
       "  handle_add_output(event, handle)\n",
       "}\n",
       "\n",
       "function register_renderer(events, OutputArea) {\n",
       "  function append_mime(data, metadata, element) {\n",
       "    // create a DOM node to render to\n",
       "    var toinsert = this.create_output_subarea(\n",
       "    metadata,\n",
       "    CLASS_NAME,\n",
       "    EXEC_MIME_TYPE\n",
       "    );\n",
       "    this.keyboard_manager.register_events(toinsert);\n",
       "    // Render to node\n",
       "    var props = {data: data, metadata: metadata[EXEC_MIME_TYPE]};\n",
       "    render(props, toinsert[0]);\n",
       "    element.append(toinsert);\n",
       "    return toinsert\n",
       "  }\n",
       "\n",
       "  events.on('output_added.OutputArea', handle_add_output);\n",
       "  events.on('output_updated.OutputArea', handle_update_output);\n",
       "  events.on('clear_output.CodeCell', handle_clear_output);\n",
       "  events.on('delete.Cell', handle_clear_output);\n",
       "  events.on('kernel_ready.Kernel', handle_kernel_cleanup);\n",
       "\n",
       "  OutputArea.prototype.register_mime_type(EXEC_MIME_TYPE, append_mime, {\n",
       "    safe: true,\n",
       "    index: 0\n",
       "  });\n",
       "}\n",
       "\n",
       "if (window.Jupyter !== undefined) {\n",
       "  try {\n",
       "    var events = require('base/js/events');\n",
       "    var OutputArea = require('notebook/js/outputarea').OutputArea;\n",
       "    if (OutputArea.prototype.mime_types().indexOf(EXEC_MIME_TYPE) == -1) {\n",
       "      register_renderer(events, OutputArea);\n",
       "    }\n",
       "  } catch(err) {\n",
       "  }\n",
       "}\n"
      ],
      "application/vnd.holoviews_load.v0+json": "\nif ((window.PyViz === undefined) || (window.PyViz instanceof HTMLElement)) {\n  window.PyViz = {comms: {}, comm_status:{}, kernels:{}, receivers: {}, plot_index: []}\n}\n\n\n    function JupyterCommManager() {\n    }\n\n    JupyterCommManager.prototype.register_target = function(plot_id, comm_id, msg_handler) {\n      if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n        comm_manager.register_target(comm_id, function(comm) {\n          comm.on_msg(msg_handler);\n        });\n      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n        window.PyViz.kernels[plot_id].registerCommTarget(comm_id, function(comm) {\n          comm.onMsg = msg_handler;\n        });\n      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n        google.colab.kernel.comms.registerTarget(comm_id, (comm) => {\n          var messages = comm.messages[Symbol.asyncIterator]();\n          function processIteratorResult(result) {\n            var message = result.value;\n            console.log(message)\n            var content = {data: message.data, comm_id};\n            var buffers = []\n            for (var buffer of message.buffers || []) {\n              buffers.push(new DataView(buffer))\n            }\n            var metadata = message.metadata || {};\n            var msg = {content, buffers, metadata}\n            msg_handler(msg);\n            return messages.next().then(processIteratorResult);\n          }\n          return messages.next().then(processIteratorResult);\n        })\n      }\n    }\n\n    JupyterCommManager.prototype.get_client_comm = function(plot_id, comm_id, msg_handler) {\n      if (comm_id in window.PyViz.comms) {\n        return window.PyViz.comms[comm_id];\n      } else if (window.comm_manager || ((window.Jupyter !== undefined) && (Jupyter.notebook.kernel != null))) {\n        var comm_manager = window.comm_manager || Jupyter.notebook.kernel.comm_manager;\n        var comm = comm_manager.new_comm(comm_id, {}, {}, {}, comm_id);\n        if (msg_handler) {\n          comm.on_msg(msg_handler);\n        }\n      } else if ((plot_id in window.PyViz.kernels) && (window.PyViz.kernels[plot_id])) {\n        var comm = window.PyViz.kernels[plot_id].connectToComm(comm_id);\n        comm.open();\n        if (msg_handler) {\n          comm.onMsg = msg_handler;\n        }\n      } else if (typeof google != 'undefined' && google.colab.kernel != null) {\n        var comm_promise = google.colab.kernel.comms.open(comm_id)\n        comm_promise.then((comm) => {\n          window.PyViz.comms[comm_id] = comm;\n          if (msg_handler) {\n            var messages = comm.messages[Symbol.asyncIterator]();\n            function processIteratorResult(result) {\n              var message = result.value;\n              var content = {data: message.data};\n              var metadata = message.metadata || {comm_id};\n              var msg = {content, metadata}\n              msg_handler(msg);\n              return messages.next().then(processIteratorResult);\n            }\n            return messages.next().then(processIteratorResult);\n          }\n        }) \n        var sendClosure = (data, metadata, buffers, disposeOnDone) => {\n          return comm_promise.then((comm) => {\n            comm.send(data, metadata, buffers, disposeOnDone);\n          });\n        };\n        var comm = {\n          send: sendClosure\n        };\n      }\n      window.PyViz.comms[comm_id] = comm;\n      return comm;\n    }\n    window.PyViz.comm_manager = new JupyterCommManager();\n    \n\n\nvar JS_MIME_TYPE = 'application/javascript';\nvar HTML_MIME_TYPE = 'text/html';\nvar EXEC_MIME_TYPE = 'application/vnd.holoviews_exec.v0+json';\nvar CLASS_NAME = 'output';\n\n/**\n * Render data to the DOM node\n */\nfunction render(props, node) {\n  var div = document.createElement(\"div\");\n  var script = document.createElement(\"script\");\n  node.appendChild(div);\n  node.appendChild(script);\n}\n\n/**\n * Handle when a new output is added\n */\nfunction handle_add_output(event, handle) {\n  var output_area = handle.output_area;\n  var output = handle.output;\n  if ((output.data == undefined) || (!output.data.hasOwnProperty(EXEC_MIME_TYPE))) {\n    return\n  }\n  var id = output.metadata[EXEC_MIME_TYPE][\"id\"];\n  var toinsert = output_area.element.find(\".\" + CLASS_NAME.split(' ')[0]);\n  if (id !== undefined) {\n    var nchildren = toinsert.length;\n    var html_node = toinsert[nchildren-1].children[0];\n    html_node.innerHTML = output.data[HTML_MIME_TYPE];\n    var scripts = [];\n    var nodelist = html_node.querySelectorAll(\"script\");\n    for (var i in nodelist) {\n      if (nodelist.hasOwnProperty(i)) {\n        scripts.push(nodelist[i])\n      }\n    }\n\n    scripts.forEach( function (oldScript) {\n      var newScript = document.createElement(\"script\");\n      var attrs = [];\n      var nodemap = oldScript.attributes;\n      for (var j in nodemap) {\n        if (nodemap.hasOwnProperty(j)) {\n          attrs.push(nodemap[j])\n        }\n      }\n      attrs.forEach(function(attr) { newScript.setAttribute(attr.name, attr.value) });\n      newScript.appendChild(document.createTextNode(oldScript.innerHTML));\n      oldScript.parentNode.replaceChild(newScript, oldScript);\n    });\n    if (JS_MIME_TYPE in output.data) {\n      toinsert[nchildren-1].children[1].textContent = output.data[JS_MIME_TYPE];\n    }\n    output_area._hv_plot_id = id;\n    if ((window.Bokeh !== undefined) && (id in Bokeh.index)) {\n      window.PyViz.plot_index[id] = Bokeh.index[id];\n    } else {\n      window.PyViz.plot_index[id] = null;\n    }\n  } else if (output.metadata[EXEC_MIME_TYPE][\"server_id\"] !== undefined) {\n    var bk_div = document.createElement(\"div\");\n    bk_div.innerHTML = output.data[HTML_MIME_TYPE];\n    var script_attrs = bk_div.children[0].attributes;\n    for (var i = 0; i < script_attrs.length; i++) {\n      toinsert[toinsert.length - 1].childNodes[1].setAttribute(script_attrs[i].name, script_attrs[i].value);\n    }\n    // store reference to server id on output_area\n    output_area._bokeh_server_id = output.metadata[EXEC_MIME_TYPE][\"server_id\"];\n  }\n}\n\n/**\n * Handle when an output is cleared or removed\n */\nfunction handle_clear_output(event, handle) {\n  var id = handle.cell.output_area._hv_plot_id;\n  var server_id = handle.cell.output_area._bokeh_server_id;\n  if (((id === undefined) || !(id in PyViz.plot_index)) && (server_id !== undefined)) { return; }\n  var comm = window.PyViz.comm_manager.get_client_comm(\"hv-extension-comm\", \"hv-extension-comm\", function () {});\n  if (server_id !== null) {\n    comm.send({event_type: 'server_delete', 'id': server_id});\n    return;\n  } else if (comm !== null) {\n    comm.send({event_type: 'delete', 'id': id});\n  }\n  delete PyViz.plot_index[id];\n  if ((window.Bokeh !== undefined) & (id in window.Bokeh.index)) {\n    var doc = window.Bokeh.index[id].model.document\n    doc.clear();\n    const i = window.Bokeh.documents.indexOf(doc);\n    if (i > -1) {\n      window.Bokeh.documents.splice(i, 1);\n    }\n  }\n}\n\n/**\n * Handle kernel restart event\n */\nfunction handle_kernel_cleanup(event, handle) {\n  delete PyViz.comms[\"hv-extension-comm\"];\n  window.PyViz.plot_index = {}\n}\n\n/**\n * Handle update_display_data messages\n */\nfunction handle_update_output(event, handle) {\n  handle_clear_output(event, {cell: {output_area: handle.output_area}})\n  handle_add_output(event, handle)\n}\n\nfunction register_renderer(events, OutputArea) {\n  function append_mime(data, metadata, element) {\n    // create a DOM node to render to\n    var toinsert = this.create_output_subarea(\n    metadata,\n    CLASS_NAME,\n    EXEC_MIME_TYPE\n    );\n    this.keyboard_manager.register_events(toinsert);\n    // Render to node\n    var props = {data: data, metadata: metadata[EXEC_MIME_TYPE]};\n    render(props, toinsert[0]);\n    element.append(toinsert);\n    return toinsert\n  }\n\n  events.on('output_added.OutputArea', handle_add_output);\n  events.on('output_updated.OutputArea', handle_update_output);\n  events.on('clear_output.CodeCell', handle_clear_output);\n  events.on('delete.Cell', handle_clear_output);\n  events.on('kernel_ready.Kernel', handle_kernel_cleanup);\n\n  OutputArea.prototype.register_mime_type(EXEC_MIME_TYPE, append_mime, {\n    safe: true,\n    index: 0\n  });\n}\n\nif (window.Jupyter !== undefined) {\n  try {\n    var events = require('base/js/events');\n    var OutputArea = require('notebook/js/outputarea').OutputArea;\n    if (OutputArea.prototype.mime_types().indexOf(EXEC_MIME_TYPE) == -1) {\n      register_renderer(events, OutputArea);\n    }\n  } catch(err) {\n  }\n}\n"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<style>.bk-root, .bk-root .bk:before, .bk-root .bk:after {\n",
       "  font-family: var(--jp-ui-font-size1);\n",
       "  font-size: var(--jp-ui-font-size1);\n",
       "  color: var(--jp-ui-font-color1);\n",
       "}\n",
       "</style>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Base imports\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from cml.plot import initialize_bokeh\n",
    "from cml.panel import initialize_panel\n",
    "from jupyterthemes.stylefx import set_nb_theme\n",
    "from bokeh.io import show\n",
    "initialize_bokeh()\n",
    "initialize_panel()\n",
    "set_nb_theme(\"onedork\")\n",
    "rng = np.random.RandomState(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"regression\"></a>\n",
    "## Simple learning problem\n",
    "\n",
    "Imagine that a certain process somewhere follows the form of a quadratic relationship\n",
    "\n",
    "$$\n",
    " y = a x^{2} + bx + c \n",
    "$$\n",
    "\n",
    "In this case, all the **unknown parameters** are that of a polynomial model, therefore we have $\\theta = \\{a, b, c\\}$. However, this is clearly an ideal (clean) case, whereas in natural observations, there might be some noise in our observations\n",
    "$$\n",
    " y = a x^{2} + bx + c +\\epsilon \\quad \\mbox{with} \\quad \\epsilon \\in [-0.1, 0.1]\n",
    "$$\n",
    "\n",
    "An example of such noisy observations for different parameters is given below"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Parameters to our function\n",
    "eps = 1\n",
    "a, b, c = 5, 2, 0\n",
    "# Generating the corresponding data\n",
    "x = np.linspace(0, 1, 100)\n",
    "poly = np.poly1d([a, b, c])\n",
    "epsilon = np.random.uniform(-eps, eps, x.shape)\n",
    "y = poly(x) + epsilon"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.holoviews_exec.v0+json": "",
      "text/html": [
       "<div id='1061'>\n",
       "  <div class=\"bk-root\" id=\"79ca14a8-cf59-4c9a-800f-7fe370ee33a2\" data-root-id=\"1061\"></div>\n",
       "</div>\n",
       "<script type=\"application/javascript\">(function(root) {\n",
       "  function embed_document(root) {\n",
       "    var docs_json = {\"2fdb55f1-6c34-4fdf-8211-756d90a5285f\":{\"defs\":[{\"extends\":null,\"module\":null,\"name\":\"ReactiveHTML1\",\"overrides\":[],\"properties\":[]},{\"extends\":null,\"module\":null,\"name\":\"FlexBox1\",\"overrides\":[],\"properties\":[{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_content\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_items\"},{\"default\":\"row\",\"kind\":null,\"name\":\"flex_direction\"},{\"default\":\"wrap\",\"kind\":null,\"name\":\"flex_wrap\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"justify_content\"}]},{\"extends\":null,\"module\":null,\"name\":\"GridStack1\",\"overrides\":[],\"properties\":[{\"default\":\"warn\",\"kind\":null,\"name\":\"mode\"},{\"default\":null,\"kind\":null,\"name\":\"ncols\"},{\"default\":null,\"kind\":null,\"name\":\"nrows\"},{\"default\":true,\"kind\":null,\"name\":\"allow_resize\"},{\"default\":true,\"kind\":null,\"name\":\"allow_drag\"},{\"default\":[],\"kind\":null,\"name\":\"state\"}]},{\"extends\":null,\"module\":null,\"name\":\"click1\",\"overrides\":[],\"properties\":[{\"default\":\"\",\"kind\":null,\"name\":\"terminal_output\"},{\"default\":\"\",\"kind\":null,\"name\":\"debug_name\"},{\"default\":0,\"kind\":null,\"name\":\"clears\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationAreaBase1\",\"overrides\":[],\"properties\":[{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationArea1\",\"overrides\":[],\"properties\":[{\"default\":[],\"kind\":null,\"name\":\"notifications\"},{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"},{\"default\":[{\"background\":\"#ffc107\",\"icon\":{\"className\":\"fas fa-exclamation-triangle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"warning\"},{\"background\":\"#007bff\",\"icon\":{\"className\":\"fas fa-info-circle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"info\"}],\"kind\":null,\"name\":\"types\"}]},{\"extends\":null,\"module\":null,\"name\":\"Notification\",\"overrides\":[],\"properties\":[{\"default\":null,\"kind\":null,\"name\":\"background\"},{\"default\":3000,\"kind\":null,\"name\":\"duration\"},{\"default\":null,\"kind\":null,\"name\":\"icon\"},{\"default\":\"\",\"kind\":null,\"name\":\"message\"},{\"default\":null,\"kind\":null,\"name\":\"notification_type\"},{\"default\":false,\"kind\":null,\"name\":\"_destroyed\"}]},{\"extends\":null,\"module\":null,\"name\":\"TemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]},{\"extends\":null,\"module\":null,\"name\":\"MaterialTemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]}],\"roots\":{\"references\":[{\"attributes\":{},\"id\":\"1012\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"1048\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00102\",\"sizing_mode\":\"stretch_width\"},\"id\":\"1085\",\"type\":\"Spacer\"},{\"attributes\":{},\"id\":\"1049\",\"type\":\"AllLabels\"},{\"attributes\":{\"reload\":false},\"id\":\"1153\",\"type\":\"panel.models.location.Location\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Simple quadratic problem\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"1004\",\"type\":\"Title\"},{\"attributes\":{},\"id\":\"1015\",\"type\":\"BasicTicker\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1058\",\"type\":\"Circle\"},{\"attributes\":{\"axis\":{\"id\":\"1018\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1021\",\"type\":\"Grid\"},{\"attributes\":{\"client_comm_id\":\"30f989f85c344654b997952eaa27007a\",\"comm_id\":\"7d56aaf9b8cc4eae98c0565cc8af0a06\",\"plot_id\":\"1061\"},\"id\":\"1119\",\"type\":\"panel.models.comm_manager.CommManager\"},{\"attributes\":{\"axis_label\":\"y\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1045\"},\"group\":null,\"major_label_policy\":{\"id\":\"1046\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1019\"}},\"id\":\"1018\",\"type\":\"LinearAxis\"},{\"attributes\":{},\"id\":\"1050\",\"type\":\"UnionRenderers\"},{\"attributes\":{},\"id\":\"1010\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"1051\",\"type\":\"Selection\"},{\"attributes\":{},\"id\":\"1027\",\"type\":\"HelpTool\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"1051\"},\"selection_policy\":{\"id\":\"1050\"}},\"id\":\"1036\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1039\",\"type\":\"Circle\"},{\"attributes\":{\"axis\":{\"id\":\"1014\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1017\",\"type\":\"Grid\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1073\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"1041\"}]},\"id\":\"1054\",\"type\":\"LegendItem\"},{\"attributes\":{},\"id\":\"1023\",\"type\":\"WheelZoomTool\"},{\"attributes\":{\"source\":{\"id\":\"1036\"}},\"id\":\"1042\",\"type\":\"CDSView\"},{\"attributes\":{},\"id\":\"1008\",\"type\":\"DataRange1d\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"1014\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"1017\"},{\"id\":\"1021\"},{\"id\":\"1053\"}],\"height\":450,\"left\":[{\"id\":\"1018\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"1041\"}],\"title\":{\"id\":\"1004\"},\"toolbar\":{\"id\":\"1029\"},\"toolbar_location\":\"left\",\"x_range\":{\"id\":\"1006\"},\"x_scale\":{\"id\":\"1010\"},\"y_range\":{\"id\":\"1008\"},\"y_scale\":{\"id\":\"1012\"}},\"id\":\"1003\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{},\"id\":\"1045\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{},\"id\":\"1019\",\"type\":\"BasicTicker\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"1073\"},\"renderers\":[{\"id\":\"1041\"}]},\"id\":\"1059\",\"type\":\"BoxSelectTool\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1028\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1037\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1006\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"1046\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"1025\",\"type\":\"SaveTool\"},{\"attributes\":{},\"id\":\"1026\",\"type\":\"ResetTool\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"1054\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"1053\",\"type\":\"Legend\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"1036\"},\"glyph\":{\"id\":\"1037\"},\"group\":null,\"hover_glyph\":{\"id\":\"1039\"},\"muted_glyph\":{\"id\":\"1040\"},\"nonselection_glyph\":{\"id\":\"1058\"},\"selection_glyph\":{\"id\":\"1057\"},\"view\":{\"id\":\"1042\"}},\"id\":\"1041\",\"type\":\"GlyphRenderer\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1040\",\"type\":\"Circle\"},{\"attributes\":{\"axis_label\":\"x\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1048\"},\"group\":null,\"major_label_policy\":{\"id\":\"1049\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1015\"}},\"id\":\"1014\",\"type\":\"LinearAxis\"},{\"attributes\":{\"children\":[{\"id\":\"1062\"},{\"id\":\"1003\"},{\"id\":\"1085\"}],\"margin\":[0,0,0,0],\"name\":\"Row00105\"},\"id\":\"1061\",\"type\":\"Row\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1057\",\"type\":\"Circle\"},{\"attributes\":{\"overlay\":{\"id\":\"1028\"}},\"id\":\"1024\",\"type\":\"BoxZoomTool\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"1041\"}],\"tooltips\":null},\"id\":\"1055\",\"type\":\"HoverTool\"},{\"attributes\":{\"tools\":[{\"id\":\"1022\"},{\"id\":\"1023\"},{\"id\":\"1024\"},{\"id\":\"1025\"},{\"id\":\"1026\"},{\"id\":\"1027\"},{\"id\":\"1055\"},{\"id\":\"1059\"}]},\"id\":\"1029\",\"type\":\"Toolbar\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00101\",\"sizing_mode\":\"stretch_width\"},\"id\":\"1062\",\"type\":\"Spacer\"},{\"attributes\":{},\"id\":\"1022\",\"type\":\"PanTool\"}],\"root_ids\":[\"1061\",\"1119\",\"1153\"]},\"title\":\"Bokeh Application\",\"version\":\"2.4.3\"}};\n",
       "    var render_items = [{\"docid\":\"2fdb55f1-6c34-4fdf-8211-756d90a5285f\",\"root_ids\":[\"1061\"],\"roots\":{\"1061\":\"79ca14a8-cf59-4c9a-800f-7fe370ee33a2\"}}];\n",
       "    root.Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "    for (const render_item of render_items) {\n",
       "      for (const root_id of render_item.root_ids) {\n",
       "\tconst id_el = document.getElementById(root_id)\n",
       "\tif (id_el.children.length && (id_el.children[0].className === 'bk-root')) {\n",
       "\t  const root_el = id_el.children[0]\n",
       "\t  root_el.id = root_el.id + '-rendered'\n",
       "\t}\n",
       "      }\n",
       "    }\n",
       "  }\n",
       "  if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "        clearInterval(timer);\n",
       "        embed_document(root);\n",
       "      } else if (document.readyState == \"complete\") {\n",
       "        attempts++;\n",
       "        if (attempts > 200) {\n",
       "          clearInterval(timer);\n",
       "          console.log(\"Bokeh: ERROR: Unable to run BokehJS code because BokehJS library is missing\");\n",
       "        }\n",
       "      }\n",
       "    }, 25, root)\n",
       "  }\n",
       "})(window);</script>"
      ],
      "text/plain": [
       "Row\n",
       "    [0] HSpacer()\n",
       "    [1] Bokeh(Figure)\n",
       "    [2] HSpacer()"
      ]
     },
     "execution_count": 5,
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {
       "id": "1061"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from cml.plot import center_plot, scatter\n",
    "plot = (center_plot(scatter(x, y, title=\"Simple quadratic problem\", toolbar_location=\"left\")))\n",
    "plot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now our main problem is that this function can follow different types of parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "params = [[5, -5, 4], [-2, 1, 0], [0.1, 1, 1]]\n",
    "# Generating the x axis\n",
    "x = np.linspace(0, 1, 100)\n",
    "plots = []\n",
    "for p in range(len(params)):\n",
    "    poly = np.poly1d(params[p])\n",
    "    epsilon = np.random.uniform(-eps, eps, x.shape)\n",
    "    y = poly(x) + epsilon\n",
    "    plots.append(scatter(x, y, title=\"Problem \"+(str(p+1))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.holoviews_exec.v0+json": "",
      "text/html": [
       "<div id='1364'>\n",
       "  <div class=\"bk-root\" id=\"4064d49e-3baa-4585-945e-4e17fb58e3fb\" data-root-id=\"1364\"></div>\n",
       "</div>\n",
       "<script type=\"application/javascript\">(function(root) {\n",
       "  function embed_document(root) {\n",
       "    var docs_json = {\"02f88535-5f68-45d6-a898-be7082dd3beb\":{\"defs\":[{\"extends\":null,\"module\":null,\"name\":\"ReactiveHTML1\",\"overrides\":[],\"properties\":[]},{\"extends\":null,\"module\":null,\"name\":\"FlexBox1\",\"overrides\":[],\"properties\":[{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_content\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_items\"},{\"default\":\"row\",\"kind\":null,\"name\":\"flex_direction\"},{\"default\":\"wrap\",\"kind\":null,\"name\":\"flex_wrap\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"justify_content\"}]},{\"extends\":null,\"module\":null,\"name\":\"GridStack1\",\"overrides\":[],\"properties\":[{\"default\":\"warn\",\"kind\":null,\"name\":\"mode\"},{\"default\":null,\"kind\":null,\"name\":\"ncols\"},{\"default\":null,\"kind\":null,\"name\":\"nrows\"},{\"default\":true,\"kind\":null,\"name\":\"allow_resize\"},{\"default\":true,\"kind\":null,\"name\":\"allow_drag\"},{\"default\":[],\"kind\":null,\"name\":\"state\"}]},{\"extends\":null,\"module\":null,\"name\":\"click1\",\"overrides\":[],\"properties\":[{\"default\":\"\",\"kind\":null,\"name\":\"terminal_output\"},{\"default\":\"\",\"kind\":null,\"name\":\"debug_name\"},{\"default\":0,\"kind\":null,\"name\":\"clears\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationAreaBase1\",\"overrides\":[],\"properties\":[{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationArea1\",\"overrides\":[],\"properties\":[{\"default\":[],\"kind\":null,\"name\":\"notifications\"},{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"},{\"default\":[{\"background\":\"#ffc107\",\"icon\":{\"className\":\"fas fa-exclamation-triangle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"warning\"},{\"background\":\"#007bff\",\"icon\":{\"className\":\"fas fa-info-circle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"info\"}],\"kind\":null,\"name\":\"types\"}]},{\"extends\":null,\"module\":null,\"name\":\"Notification\",\"overrides\":[],\"properties\":[{\"default\":null,\"kind\":null,\"name\":\"background\"},{\"default\":3000,\"kind\":null,\"name\":\"duration\"},{\"default\":null,\"kind\":null,\"name\":\"icon\"},{\"default\":\"\",\"kind\":null,\"name\":\"message\"},{\"default\":null,\"kind\":null,\"name\":\"notification_type\"},{\"default\":false,\"kind\":null,\"name\":\"_destroyed\"}]},{\"extends\":null,\"module\":null,\"name\":\"TemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]},{\"extends\":null,\"module\":null,\"name\":\"MaterialTemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]}],\"roots\":{\"references\":[{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00109\",\"sizing_mode\":\"stretch_width\"},\"id\":\"1365\",\"type\":\"Spacer\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"1354\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"1353\",\"type\":\"Legend\"},{\"attributes\":{\"source\":{\"id\":\"1336\"}},\"id\":\"1342\",\"type\":\"CDSView\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00110\",\"sizing_mode\":\"stretch_width\"},\"id\":\"1432\",\"type\":\"Spacer\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1282\",\"type\":\"Circle\"},{\"attributes\":{\"children\":[{\"id\":\"1365\"},{\"id\":\"1363\"},{\"id\":\"1432\"}],\"margin\":[0,0,0,0],\"name\":\"Row00113\"},\"id\":\"1364\",\"type\":\"Row\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"1238\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"1237\",\"type\":\"Legend\"},{\"attributes\":{},\"id\":\"1209\",\"type\":\"SaveTool\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1281\",\"type\":\"Circle\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"1336\"},\"glyph\":{\"id\":\"1337\"},\"group\":null,\"hover_glyph\":{\"id\":\"1339\"},\"muted_glyph\":{\"id\":\"1340\"},\"nonselection_glyph\":{\"id\":\"1358\"},\"selection_glyph\":{\"id\":\"1357\"},\"view\":{\"id\":\"1342\"}},\"id\":\"1341\",\"type\":\"GlyphRenderer\"},{\"attributes\":{},\"id\":\"1206\",\"type\":\"PanTool\"},{\"attributes\":{\"source\":{\"id\":\"1278\"}},\"id\":\"1284\",\"type\":\"CDSView\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1224\",\"type\":\"Circle\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"1283\"}]},\"id\":\"1296\",\"type\":\"LegendItem\"},{\"attributes\":{\"source\":{\"id\":\"1220\"}},\"id\":\"1226\",\"type\":\"CDSView\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"1225\"}]},\"id\":\"1238\",\"type\":\"LegendItem\"},{\"attributes\":{\"children\":[{\"id\":\"1361\"},{\"id\":\"1362\"}]},\"id\":\"1363\",\"type\":\"Column\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"1296\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"1295\",\"type\":\"Legend\"},{\"attributes\":{},\"id\":\"1203\",\"type\":\"BasicTicker\"},{\"attributes\":{\"overlay\":{\"id\":\"1212\"}},\"id\":\"1208\",\"type\":\"BoxZoomTool\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Problem 1\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"1188\",\"type\":\"Title\"},{\"attributes\":{},\"id\":\"1207\",\"type\":\"WheelZoomTool\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1398\",\"type\":\"BoxAnnotation\"},{\"attributes\":{},\"id\":\"1194\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"1190\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"1292\",\"type\":\"UnionRenderers\"},{\"attributes\":{},\"id\":\"1351\",\"type\":\"Selection\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"1198\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"1201\"},{\"id\":\"1205\"},{\"id\":\"1237\"}],\"height\":450,\"left\":[{\"id\":\"1202\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"1225\"}],\"title\":{\"id\":\"1188\"},\"toolbar\":{\"id\":\"1213\"},\"toolbar_location\":null,\"x_range\":{\"id\":\"1190\"},\"x_scale\":{\"id\":\"1194\"},\"y_range\":{\"id\":\"1192\"},\"y_scale\":{\"id\":\"1196\"}},\"id\":\"1187\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{},\"id\":\"1291\",\"type\":\"AllLabels\"},{\"attributes\":{\"tools\":[{\"id\":\"1206\"},{\"id\":\"1207\"},{\"id\":\"1208\"},{\"id\":\"1209\"},{\"id\":\"1210\"},{\"id\":\"1211\"},{\"id\":\"1239\"},{\"id\":\"1243\"}]},\"id\":\"1213\",\"type\":\"Toolbar\"},{\"attributes\":{},\"id\":\"1346\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"1229\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1394\",\"type\":\"BoxAnnotation\"},{\"attributes\":{},\"id\":\"1287\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1357\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1348\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{},\"id\":\"1230\",\"type\":\"AllLabels\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"1398\"},\"renderers\":[{\"id\":\"1341\"}]},\"id\":\"1359\",\"type\":\"BoxSelectTool\"},{\"attributes\":{\"axis\":{\"id\":\"1198\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1201\",\"type\":\"Grid\"},{\"attributes\":{},\"id\":\"1288\",\"type\":\"AllLabels\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1340\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1234\",\"type\":\"UnionRenderers\"},{\"attributes\":{},\"id\":\"1345\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1299\",\"type\":\"Circle\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"1293\"},\"selection_policy\":{\"id\":\"1292\"}},\"id\":\"1278\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"1341\"}],\"tooltips\":null},\"id\":\"1355\",\"type\":\"HoverTool\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"1394\"},\"renderers\":[{\"id\":\"1283\"}]},\"id\":\"1301\",\"type\":\"BoxSelectTool\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1241\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1293\",\"type\":\"Selection\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"1390\"},\"renderers\":[{\"id\":\"1225\"}]},\"id\":\"1243\",\"type\":\"BoxSelectTool\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"1235\"},\"selection_policy\":{\"id\":\"1234\"}},\"id\":\"1220\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"1283\"}],\"tooltips\":null},\"id\":\"1297\",\"type\":\"HoverTool\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1358\",\"type\":\"Circle\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"1314\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"1317\"},{\"id\":\"1321\"},{\"id\":\"1353\"}],\"height\":450,\"left\":[{\"id\":\"1318\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"1341\"}],\"title\":{\"id\":\"1304\"},\"toolbar\":{\"id\":\"1329\"},\"toolbar_location\":null,\"x_range\":{\"id\":\"1306\"},\"x_scale\":{\"id\":\"1310\"},\"y_range\":{\"id\":\"1308\"},\"y_scale\":{\"id\":\"1312\"}},\"id\":\"1303\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{\"axis_label\":\"x\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1232\"},\"group\":null,\"major_label_policy\":{\"id\":\"1233\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1199\"}},\"id\":\"1198\",\"type\":\"LinearAxis\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"1225\"}],\"tooltips\":null},\"id\":\"1239\",\"type\":\"HoverTool\"},{\"attributes\":{\"axis\":{\"id\":\"1202\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1205\",\"type\":\"Grid\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"1256\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"1259\"},{\"id\":\"1263\"},{\"id\":\"1295\"}],\"height\":450,\"left\":[{\"id\":\"1260\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"1283\"}],\"title\":{\"id\":\"1246\"},\"toolbar\":{\"id\":\"1271\"},\"toolbar_location\":null,\"x_range\":{\"id\":\"1248\"},\"x_scale\":{\"id\":\"1252\"},\"y_range\":{\"id\":\"1250\"},\"y_scale\":{\"id\":\"1254\"}},\"id\":\"1245\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{},\"id\":\"1210\",\"type\":\"ResetTool\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1337\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1235\",\"type\":\"Selection\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1300\",\"type\":\"Circle\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"1278\"},\"glyph\":{\"id\":\"1279\"},\"group\":null,\"hover_glyph\":{\"id\":\"1281\"},\"muted_glyph\":{\"id\":\"1282\"},\"nonselection_glyph\":{\"id\":\"1300\"},\"selection_glyph\":{\"id\":\"1299\"},\"view\":{\"id\":\"1284\"}},\"id\":\"1283\",\"type\":\"GlyphRenderer\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1242\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1306\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"1308\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"1199\",\"type\":\"BasicTicker\"},{\"attributes\":{},\"id\":\"1248\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"1310\",\"type\":\"LinearScale\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1212\",\"type\":\"BoxAnnotation\"},{\"attributes\":{},\"id\":\"1250\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"1350\",\"type\":\"UnionRenderers\"},{\"attributes\":{},\"id\":\"1327\",\"type\":\"HelpTool\"},{\"attributes\":{},\"id\":\"1252\",\"type\":\"LinearScale\"},{\"attributes\":{\"client_comm_id\":\"37a3aef46f12475aa16117b113224710\",\"comm_id\":\"ea061abe275247ad9a1bf58cd8564bef\",\"plot_id\":\"1364\"},\"id\":\"1532\",\"type\":\"panel.models.comm_manager.CommManager\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"1220\"},\"glyph\":{\"id\":\"1221\"},\"group\":null,\"hover_glyph\":{\"id\":\"1223\"},\"muted_glyph\":{\"id\":\"1224\"},\"nonselection_glyph\":{\"id\":\"1242\"},\"selection_glyph\":{\"id\":\"1241\"},\"view\":{\"id\":\"1226\"}},\"id\":\"1225\",\"type\":\"GlyphRenderer\"},{\"attributes\":{},\"id\":\"1269\",\"type\":\"HelpTool\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"1341\"}]},\"id\":\"1354\",\"type\":\"LegendItem\"},{\"attributes\":{},\"id\":\"1315\",\"type\":\"BasicTicker\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Problem 3\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"1304\",\"type\":\"Title\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1223\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1257\",\"type\":\"BasicTicker\"},{\"attributes\":{\"axis_label\":\"x\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1348\"},\"group\":null,\"major_label_policy\":{\"id\":\"1349\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1315\"}},\"id\":\"1314\",\"type\":\"LinearAxis\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Problem 2\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"1246\",\"type\":\"Title\"},{\"attributes\":{\"axis_label\":\"y\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1345\"},\"group\":null,\"major_label_policy\":{\"id\":\"1346\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1319\"}},\"id\":\"1318\",\"type\":\"LinearAxis\"},{\"attributes\":{\"axis_label\":\"x\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1290\"},\"group\":null,\"major_label_policy\":{\"id\":\"1291\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1257\"}},\"id\":\"1256\",\"type\":\"LinearAxis\"},{\"attributes\":{},\"id\":\"1312\",\"type\":\"LinearScale\"},{\"attributes\":{\"axis_label\":\"y\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1287\"},\"group\":null,\"major_label_policy\":{\"id\":\"1288\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1261\"}},\"id\":\"1260\",\"type\":\"LinearAxis\"},{\"attributes\":{\"children\":[{\"id\":\"1187\"},{\"id\":\"1245\"},{\"id\":\"1303\"}]},\"id\":\"1362\",\"type\":\"Row\"},{\"attributes\":{\"axis\":{\"id\":\"1314\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1317\",\"type\":\"Grid\"},{\"attributes\":{},\"id\":\"1254\",\"type\":\"LinearScale\"},{\"attributes\":{\"axis\":{\"id\":\"1256\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1259\",\"type\":\"Grid\"},{\"attributes\":{\"style\":{\"font-size\":\"250%\"},\"text\":\"Observing different problems\"},\"id\":\"1361\",\"type\":\"Div\"},{\"attributes\":{\"axis\":{\"id\":\"1318\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1321\",\"type\":\"Grid\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1279\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1211\",\"type\":\"HelpTool\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"1351\"},\"selection_policy\":{\"id\":\"1350\"}},\"id\":\"1336\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"axis\":{\"id\":\"1260\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1263\",\"type\":\"Grid\"},{\"attributes\":{},\"id\":\"1319\",\"type\":\"BasicTicker\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1339\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1261\",\"type\":\"BasicTicker\"},{\"attributes\":{},\"id\":\"1323\",\"type\":\"WheelZoomTool\"},{\"attributes\":{},\"id\":\"1322\",\"type\":\"PanTool\"},{\"attributes\":{},\"id\":\"1265\",\"type\":\"WheelZoomTool\"},{\"attributes\":{\"axis_label\":\"y\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1229\"},\"group\":null,\"major_label_policy\":{\"id\":\"1230\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1203\"}},\"id\":\"1202\",\"type\":\"LinearAxis\"},{\"attributes\":{\"overlay\":{\"id\":\"1328\"}},\"id\":\"1324\",\"type\":\"BoxZoomTool\"},{\"attributes\":{},\"id\":\"1192\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"1233\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"1264\",\"type\":\"PanTool\"},{\"attributes\":{\"overlay\":{\"id\":\"1270\"}},\"id\":\"1266\",\"type\":\"BoxZoomTool\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1390\",\"type\":\"BoxAnnotation\"},{\"attributes\":{},\"id\":\"1325\",\"type\":\"SaveTool\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1221\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1267\",\"type\":\"SaveTool\"},{\"attributes\":{},\"id\":\"1326\",\"type\":\"ResetTool\"},{\"attributes\":{},\"id\":\"1232\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{},\"id\":\"1268\",\"type\":\"ResetTool\"},{\"attributes\":{},\"id\":\"1349\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"1196\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"1290\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1328\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1270\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"tools\":[{\"id\":\"1322\"},{\"id\":\"1323\"},{\"id\":\"1324\"},{\"id\":\"1325\"},{\"id\":\"1326\"},{\"id\":\"1327\"},{\"id\":\"1355\"},{\"id\":\"1359\"}]},\"id\":\"1329\",\"type\":\"Toolbar\"},{\"attributes\":{\"tools\":[{\"id\":\"1264\"},{\"id\":\"1265\"},{\"id\":\"1266\"},{\"id\":\"1267\"},{\"id\":\"1268\"},{\"id\":\"1269\"},{\"id\":\"1297\"},{\"id\":\"1301\"}]},\"id\":\"1271\",\"type\":\"Toolbar\"}],\"root_ids\":[\"1364\",\"1532\"]},\"title\":\"Bokeh Application\",\"version\":\"2.4.3\"}};\n",
       "    var render_items = [{\"docid\":\"02f88535-5f68-45d6-a898-be7082dd3beb\",\"root_ids\":[\"1364\"],\"roots\":{\"1364\":\"4064d49e-3baa-4585-945e-4e17fb58e3fb\"}}];\n",
       "    root.Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "    for (const render_item of render_items) {\n",
       "      for (const root_id of render_item.root_ids) {\n",
       "\tconst id_el = document.getElementById(root_id)\n",
       "\tif (id_el.children.length && (id_el.children[0].className === 'bk-root')) {\n",
       "\t  const root_el = id_el.children[0]\n",
       "\t  root_el.id = root_el.id + '-rendered'\n",
       "\t}\n",
       "      }\n",
       "    }\n",
       "  }\n",
       "  if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "        clearInterval(timer);\n",
       "        embed_document(root);\n",
       "      } else if (document.readyState == \"complete\") {\n",
       "        attempts++;\n",
       "        if (attempts > 200) {\n",
       "          clearInterval(timer);\n",
       "          console.log(\"Bokeh: ERROR: Unable to run BokehJS code because BokehJS library is missing\");\n",
       "        }\n",
       "      }\n",
       "    }, 25, root)\n",
       "  }\n",
       "})(window);</script>"
      ],
      "text/plain": [
       "Row\n",
       "    [0] HSpacer()\n",
       "    [1] Bokeh(Column)\n",
       "    [2] HSpacer()"
      ]
     },
     "execution_count": 7,
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {
       "id": "1364"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from bokeh.models import Div\n",
    "from bokeh.layouts import column, row\n",
    "plot = center_plot(column(Div(text = \"Observing different problems\", style={'font-size': '250%'}), row(*plots)))\n",
    "plot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In real-life settings, this function can also have different levels of noise, as exemplified in the following code."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "params = [3, 0, 1]\n",
    "noise_levels = [0.1, 1.0, 8.0]\n",
    "# Generating the x axis\n",
    "x = np.linspace(0, 1, 100)\n",
    "plots = []\n",
    "for p in range(len(noise_levels)):\n",
    "    poly = np.poly1d(params)\n",
    "    epsilon = np.random.uniform(-noise_levels[p],noise_levels[p],x.shape)\n",
    "    y = poly(x) + epsilon\n",
    "    plots.append(scatter(x, y, title=\"Problem \"+(str(p+1))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.holoviews_exec.v0+json": "",
      "text/html": [
       "<div id='1875'>\n",
       "  <div class=\"bk-root\" id=\"53d5e025-dba5-4c9d-b81d-59e2834895ec\" data-root-id=\"1875\"></div>\n",
       "</div>\n",
       "<script type=\"application/javascript\">(function(root) {\n",
       "  function embed_document(root) {\n",
       "    var docs_json = {\"2230d6e1-153f-4d54-9329-82656ae3f26f\":{\"defs\":[{\"extends\":null,\"module\":null,\"name\":\"ReactiveHTML1\",\"overrides\":[],\"properties\":[]},{\"extends\":null,\"module\":null,\"name\":\"FlexBox1\",\"overrides\":[],\"properties\":[{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_content\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_items\"},{\"default\":\"row\",\"kind\":null,\"name\":\"flex_direction\"},{\"default\":\"wrap\",\"kind\":null,\"name\":\"flex_wrap\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"justify_content\"}]},{\"extends\":null,\"module\":null,\"name\":\"GridStack1\",\"overrides\":[],\"properties\":[{\"default\":\"warn\",\"kind\":null,\"name\":\"mode\"},{\"default\":null,\"kind\":null,\"name\":\"ncols\"},{\"default\":null,\"kind\":null,\"name\":\"nrows\"},{\"default\":true,\"kind\":null,\"name\":\"allow_resize\"},{\"default\":true,\"kind\":null,\"name\":\"allow_drag\"},{\"default\":[],\"kind\":null,\"name\":\"state\"}]},{\"extends\":null,\"module\":null,\"name\":\"click1\",\"overrides\":[],\"properties\":[{\"default\":\"\",\"kind\":null,\"name\":\"terminal_output\"},{\"default\":\"\",\"kind\":null,\"name\":\"debug_name\"},{\"default\":0,\"kind\":null,\"name\":\"clears\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationAreaBase1\",\"overrides\":[],\"properties\":[{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationArea1\",\"overrides\":[],\"properties\":[{\"default\":[],\"kind\":null,\"name\":\"notifications\"},{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"},{\"default\":[{\"background\":\"#ffc107\",\"icon\":{\"className\":\"fas fa-exclamation-triangle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"warning\"},{\"background\":\"#007bff\",\"icon\":{\"className\":\"fas fa-info-circle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"info\"}],\"kind\":null,\"name\":\"types\"}]},{\"extends\":null,\"module\":null,\"name\":\"Notification\",\"overrides\":[],\"properties\":[{\"default\":null,\"kind\":null,\"name\":\"background\"},{\"default\":3000,\"kind\":null,\"name\":\"duration\"},{\"default\":null,\"kind\":null,\"name\":\"icon\"},{\"default\":\"\",\"kind\":null,\"name\":\"message\"},{\"default\":null,\"kind\":null,\"name\":\"notification_type\"},{\"default\":false,\"kind\":null,\"name\":\"_destroyed\"}]},{\"extends\":null,\"module\":null,\"name\":\"TemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]},{\"extends\":null,\"module\":null,\"name\":\"MaterialTemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]}],\"roots\":{\"references\":[{\"attributes\":{\"client_comm_id\":\"06e20ef4e0ad4a03996166d77ec26d75\",\"comm_id\":\"82b5ef88810c4db181ecf0b53a197ca4\",\"plot_id\":\"1875\"},\"id\":\"2043\",\"type\":\"panel.models.comm_manager.CommManager\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00116\",\"sizing_mode\":\"stretch_width\"},\"id\":\"1876\",\"type\":\"Spacer\"},{\"attributes\":{\"children\":[{\"id\":\"1872\"},{\"id\":\"1873\"}]},\"id\":\"1874\",\"type\":\"Column\"},{\"attributes\":{\"style\":{\"font-size\":\"250%\"},\"text\":\"Different amounts of noise\"},\"id\":\"1872\",\"type\":\"Div\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1734\",\"type\":\"Circle\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1868\",\"type\":\"Circle\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1793\",\"type\":\"Circle\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"1847\"},\"glyph\":{\"id\":\"1848\"},\"group\":null,\"hover_glyph\":{\"id\":\"1850\"},\"muted_glyph\":{\"id\":\"1851\"},\"nonselection_glyph\":{\"id\":\"1869\"},\"selection_glyph\":{\"id\":\"1868\"},\"view\":{\"id\":\"1853\"}},\"id\":\"1852\",\"type\":\"GlyphRenderer\"},{\"attributes\":{\"tools\":[{\"id\":\"1717\"},{\"id\":\"1718\"},{\"id\":\"1719\"},{\"id\":\"1720\"},{\"id\":\"1721\"},{\"id\":\"1722\"},{\"id\":\"1750\"},{\"id\":\"1754\"}]},\"id\":\"1724\",\"type\":\"Toolbar\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"1862\"},\"selection_policy\":{\"id\":\"1861\"}},\"id\":\"1847\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1723\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1839\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1909\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"1901\"},\"renderers\":[{\"id\":\"1736\"}]},\"id\":\"1754\",\"type\":\"BoxSelectTool\"},{\"attributes\":{\"tools\":[{\"id\":\"1833\"},{\"id\":\"1834\"},{\"id\":\"1835\"},{\"id\":\"1836\"},{\"id\":\"1837\"},{\"id\":\"1838\"},{\"id\":\"1866\"},{\"id\":\"1870\"}]},\"id\":\"1840\",\"type\":\"Toolbar\"},{\"attributes\":{\"source\":{\"id\":\"1731\"}},\"id\":\"1737\",\"type\":\"CDSView\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"1749\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"1748\",\"type\":\"Legend\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1851\",\"type\":\"Circle\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"1736\"}]},\"id\":\"1749\",\"type\":\"LegendItem\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1850\",\"type\":\"Circle\"},{\"attributes\":{\"children\":[{\"id\":\"1876\"},{\"id\":\"1874\"},{\"id\":\"1943\"}],\"margin\":[0,0,0,0],\"name\":\"Row00120\"},\"id\":\"1875\",\"type\":\"Row\"},{\"attributes\":{\"children\":[{\"id\":\"1698\"},{\"id\":\"1756\"},{\"id\":\"1814\"}]},\"id\":\"1873\",\"type\":\"Row\"},{\"attributes\":{\"source\":{\"id\":\"1847\"}},\"id\":\"1853\",\"type\":\"CDSView\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"1865\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"1864\",\"type\":\"Legend\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"1852\"}]},\"id\":\"1865\",\"type\":\"LegendItem\"},{\"attributes\":{},\"id\":\"1717\",\"type\":\"PanTool\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"1789\"},\"glyph\":{\"id\":\"1790\"},\"group\":null,\"hover_glyph\":{\"id\":\"1792\"},\"muted_glyph\":{\"id\":\"1793\"},\"nonselection_glyph\":{\"id\":\"1811\"},\"selection_glyph\":{\"id\":\"1810\"},\"view\":{\"id\":\"1795\"}},\"id\":\"1794\",\"type\":\"GlyphRenderer\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"1909\"},\"renderers\":[{\"id\":\"1852\"}]},\"id\":\"1870\",\"type\":\"BoxSelectTool\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1752\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1798\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1810\",\"type\":\"Circle\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"1905\"},\"renderers\":[{\"id\":\"1794\"}]},\"id\":\"1812\",\"type\":\"BoxSelectTool\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"1767\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"1770\"},{\"id\":\"1774\"},{\"id\":\"1806\"}],\"height\":450,\"left\":[{\"id\":\"1771\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"1794\"}],\"title\":{\"id\":\"1757\"},\"toolbar\":{\"id\":\"1782\"},\"toolbar_location\":null,\"x_range\":{\"id\":\"1759\"},\"x_scale\":{\"id\":\"1763\"},\"y_range\":{\"id\":\"1761\"},\"y_scale\":{\"id\":\"1765\"}},\"id\":\"1756\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{},\"id\":\"1799\",\"type\":\"AllLabels\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1848\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1722\",\"type\":\"HelpTool\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"1794\"}],\"tooltips\":null},\"id\":\"1808\",\"type\":\"HoverTool\"},{\"attributes\":{\"axis\":{\"id\":\"1709\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1712\",\"type\":\"Grid\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1753\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1859\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{},\"id\":\"1759\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"1801\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{},\"id\":\"1761\",\"type\":\"DataRange1d\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1811\",\"type\":\"Circle\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"1852\"}],\"tooltips\":null},\"id\":\"1866\",\"type\":\"HoverTool\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Problem 1\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"1699\",\"type\":\"Title\"},{\"attributes\":{},\"id\":\"1763\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"1721\",\"type\":\"ResetTool\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"1825\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"1828\"},{\"id\":\"1832\"},{\"id\":\"1864\"}],\"height\":450,\"left\":[{\"id\":\"1829\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"1852\"}],\"title\":{\"id\":\"1815\"},\"toolbar\":{\"id\":\"1840\"},\"toolbar_location\":null,\"x_range\":{\"id\":\"1817\"},\"x_scale\":{\"id\":\"1821\"},\"y_range\":{\"id\":\"1819\"},\"y_scale\":{\"id\":\"1823\"}},\"id\":\"1814\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{},\"id\":\"1780\",\"type\":\"HelpTool\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"1869\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1817\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"1819\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"1802\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"1860\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"1821\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"1768\",\"type\":\"BasicTicker\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Problem 2\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"1757\",\"type\":\"Title\"},{\"attributes\":{},\"id\":\"1838\",\"type\":\"HelpTool\"},{\"attributes\":{\"axis_label\":\"x\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1801\"},\"group\":null,\"major_label_policy\":{\"id\":\"1802\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1768\"}},\"id\":\"1767\",\"type\":\"LinearAxis\"},{\"attributes\":{\"axis_label\":\"y\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1798\"},\"group\":null,\"major_label_policy\":{\"id\":\"1799\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1772\"}},\"id\":\"1771\",\"type\":\"LinearAxis\"},{\"attributes\":{},\"id\":\"1765\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"1826\",\"type\":\"BasicTicker\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1732\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1703\",\"type\":\"DataRange1d\"},{\"attributes\":{\"axis\":{\"id\":\"1767\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1770\",\"type\":\"Grid\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Problem 3\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"1815\",\"type\":\"Title\"},{\"attributes\":{\"axis_label\":\"x\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1859\"},\"group\":null,\"major_label_policy\":{\"id\":\"1860\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1826\"}},\"id\":\"1825\",\"type\":\"LinearAxis\"},{\"attributes\":{\"axis\":{\"id\":\"1771\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1774\",\"type\":\"Grid\"},{\"attributes\":{\"axis_label\":\"y\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1856\"},\"group\":null,\"major_label_policy\":{\"id\":\"1857\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1830\"}},\"id\":\"1829\",\"type\":\"LinearAxis\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1790\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"1823\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"1772\",\"type\":\"BasicTicker\"},{\"attributes\":{},\"id\":\"1745\",\"type\":\"UnionRenderers\"},{\"attributes\":{},\"id\":\"1701\",\"type\":\"DataRange1d\"},{\"attributes\":{\"axis\":{\"id\":\"1825\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1828\",\"type\":\"Grid\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"1804\"},\"selection_policy\":{\"id\":\"1803\"}},\"id\":\"1789\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1901\",\"type\":\"BoxAnnotation\"},{\"attributes\":{},\"id\":\"1804\",\"type\":\"Selection\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"1709\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"1712\"},{\"id\":\"1716\"},{\"id\":\"1748\"}],\"height\":450,\"left\":[{\"id\":\"1713\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"1736\"}],\"title\":{\"id\":\"1699\"},\"toolbar\":{\"id\":\"1724\"},\"toolbar_location\":null,\"x_range\":{\"id\":\"1701\"},\"x_scale\":{\"id\":\"1705\"},\"y_range\":{\"id\":\"1703\"},\"y_scale\":{\"id\":\"1707\"}},\"id\":\"1698\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{\"axis\":{\"id\":\"1829\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1832\",\"type\":\"Grid\"},{\"attributes\":{},\"id\":\"1776\",\"type\":\"WheelZoomTool\"},{\"attributes\":{},\"id\":\"1775\",\"type\":\"PanTool\"},{\"attributes\":{},\"id\":\"1830\",\"type\":\"BasicTicker\"},{\"attributes\":{\"axis\":{\"id\":\"1713\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"1716\",\"type\":\"Grid\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"1746\"},\"selection_policy\":{\"id\":\"1745\"}},\"id\":\"1731\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"overlay\":{\"id\":\"1781\"}},\"id\":\"1777\",\"type\":\"BoxZoomTool\"},{\"attributes\":{},\"id\":\"1803\",\"type\":\"UnionRenderers\"},{\"attributes\":{},\"id\":\"1778\",\"type\":\"SaveTool\"},{\"attributes\":{},\"id\":\"1861\",\"type\":\"UnionRenderers\"},{\"attributes\":{},\"id\":\"1744\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"1779\",\"type\":\"ResetTool\"},{\"attributes\":{},\"id\":\"1834\",\"type\":\"WheelZoomTool\"},{\"attributes\":{},\"id\":\"1833\",\"type\":\"PanTool\"},{\"attributes\":{},\"id\":\"1857\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"1720\",\"type\":\"SaveTool\"},{\"attributes\":{},\"id\":\"1714\",\"type\":\"BasicTicker\"},{\"attributes\":{\"overlay\":{\"id\":\"1839\"}},\"id\":\"1835\",\"type\":\"BoxZoomTool\"},{\"attributes\":{},\"id\":\"1743\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{},\"id\":\"1836\",\"type\":\"SaveTool\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1781\",\"type\":\"BoxAnnotation\"},{\"attributes\":{},\"id\":\"1707\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"1837\",\"type\":\"ResetTool\"},{\"attributes\":{},\"id\":\"1856\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"tools\":[{\"id\":\"1775\"},{\"id\":\"1776\"},{\"id\":\"1777\"},{\"id\":\"1778\"},{\"id\":\"1779\"},{\"id\":\"1780\"},{\"id\":\"1808\"},{\"id\":\"1812\"}]},\"id\":\"1782\",\"type\":\"Toolbar\"},{\"attributes\":{},\"id\":\"1746\",\"type\":\"Selection\"},{\"attributes\":{},\"id\":\"1718\",\"type\":\"WheelZoomTool\"},{\"attributes\":{},\"id\":\"1710\",\"type\":\"BasicTicker\"},{\"attributes\":{},\"id\":\"1862\",\"type\":\"Selection\"},{\"attributes\":{},\"id\":\"1740\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"axis_label\":\"x\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1743\"},\"group\":null,\"major_label_policy\":{\"id\":\"1744\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1710\"}},\"id\":\"1709\",\"type\":\"LinearAxis\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1792\",\"type\":\"Circle\"},{\"attributes\":{\"overlay\":{\"id\":\"1723\"}},\"id\":\"1719\",\"type\":\"BoxZoomTool\"},{\"attributes\":{},\"id\":\"1741\",\"type\":\"AllLabels\"},{\"attributes\":{\"source\":{\"id\":\"1789\"}},\"id\":\"1795\",\"type\":\"CDSView\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00117\",\"sizing_mode\":\"stretch_width\"},\"id\":\"1943\",\"type\":\"Spacer\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"1736\"}],\"tooltips\":null},\"id\":\"1750\",\"type\":\"HoverTool\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"1807\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"1806\",\"type\":\"Legend\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"1905\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"1731\"},\"glyph\":{\"id\":\"1732\"},\"group\":null,\"hover_glyph\":{\"id\":\"1734\"},\"muted_glyph\":{\"id\":\"1735\"},\"nonselection_glyph\":{\"id\":\"1753\"},\"selection_glyph\":{\"id\":\"1752\"},\"view\":{\"id\":\"1737\"}},\"id\":\"1736\",\"type\":\"GlyphRenderer\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"1794\"}]},\"id\":\"1807\",\"type\":\"LegendItem\"},{\"attributes\":{},\"id\":\"1705\",\"type\":\"LinearScale\"},{\"attributes\":{\"axis_label\":\"y\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"1740\"},\"group\":null,\"major_label_policy\":{\"id\":\"1741\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"1714\"}},\"id\":\"1713\",\"type\":\"LinearAxis\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"1735\",\"type\":\"Circle\"}],\"root_ids\":[\"1875\",\"2043\"]},\"title\":\"Bokeh Application\",\"version\":\"2.4.3\"}};\n",
       "    var render_items = [{\"docid\":\"2230d6e1-153f-4d54-9329-82656ae3f26f\",\"root_ids\":[\"1875\"],\"roots\":{\"1875\":\"53d5e025-dba5-4c9d-b81d-59e2834895ec\"}}];\n",
       "    root.Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "    for (const render_item of render_items) {\n",
       "      for (const root_id of render_item.root_ids) {\n",
       "\tconst id_el = document.getElementById(root_id)\n",
       "\tif (id_el.children.length && (id_el.children[0].className === 'bk-root')) {\n",
       "\t  const root_el = id_el.children[0]\n",
       "\t  root_el.id = root_el.id + '-rendered'\n",
       "\t}\n",
       "      }\n",
       "    }\n",
       "  }\n",
       "  if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "        clearInterval(timer);\n",
       "        embed_document(root);\n",
       "      } else if (document.readyState == \"complete\") {\n",
       "        attempts++;\n",
       "        if (attempts > 200) {\n",
       "          clearInterval(timer);\n",
       "          console.log(\"Bokeh: ERROR: Unable to run BokehJS code because BokehJS library is missing\");\n",
       "        }\n",
       "      }\n",
       "    }, 25, root)\n",
       "  }\n",
       "})(window);</script>"
      ],
      "text/plain": [
       "Row\n",
       "    [0] HSpacer()\n",
       "    [1] Bokeh(Column)\n",
       "    [2] HSpacer()"
      ]
     },
     "execution_count": 9,
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {
       "id": "1875"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "plot = center_plot(column(Div(text = \"Different amounts of noise\", style={'font-size': '250%'}), row(*plots)))\n",
    "plot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To summarize, we will have some observations of a function, and we would like to optimize a function that gets as close as possible to the real function that generated this data. Here, we plot the real function and also _subsample_ our number of observations (having only a few points to find the corresponding function)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.holoviews_exec.v0+json": "",
      "text/html": [
       "<div id='2287'>\n",
       "  <div class=\"bk-root\" id=\"c0f2d06d-b2e2-4cc6-98d4-85472d4aa2da\" data-root-id=\"2287\"></div>\n",
       "</div>\n",
       "<script type=\"application/javascript\">(function(root) {\n",
       "  function embed_document(root) {\n",
       "    var docs_json = {\"d8321e7f-68be-4e59-83ca-6a297eb77df9\":{\"defs\":[{\"extends\":null,\"module\":null,\"name\":\"ReactiveHTML1\",\"overrides\":[],\"properties\":[]},{\"extends\":null,\"module\":null,\"name\":\"FlexBox1\",\"overrides\":[],\"properties\":[{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_content\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_items\"},{\"default\":\"row\",\"kind\":null,\"name\":\"flex_direction\"},{\"default\":\"wrap\",\"kind\":null,\"name\":\"flex_wrap\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"justify_content\"}]},{\"extends\":null,\"module\":null,\"name\":\"GridStack1\",\"overrides\":[],\"properties\":[{\"default\":\"warn\",\"kind\":null,\"name\":\"mode\"},{\"default\":null,\"kind\":null,\"name\":\"ncols\"},{\"default\":null,\"kind\":null,\"name\":\"nrows\"},{\"default\":true,\"kind\":null,\"name\":\"allow_resize\"},{\"default\":true,\"kind\":null,\"name\":\"allow_drag\"},{\"default\":[],\"kind\":null,\"name\":\"state\"}]},{\"extends\":null,\"module\":null,\"name\":\"click1\",\"overrides\":[],\"properties\":[{\"default\":\"\",\"kind\":null,\"name\":\"terminal_output\"},{\"default\":\"\",\"kind\":null,\"name\":\"debug_name\"},{\"default\":0,\"kind\":null,\"name\":\"clears\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationAreaBase1\",\"overrides\":[],\"properties\":[{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationArea1\",\"overrides\":[],\"properties\":[{\"default\":[],\"kind\":null,\"name\":\"notifications\"},{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"},{\"default\":[{\"background\":\"#ffc107\",\"icon\":{\"className\":\"fas fa-exclamation-triangle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"warning\"},{\"background\":\"#007bff\",\"icon\":{\"className\":\"fas fa-info-circle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"info\"}],\"kind\":null,\"name\":\"types\"}]},{\"extends\":null,\"module\":null,\"name\":\"Notification\",\"overrides\":[],\"properties\":[{\"default\":null,\"kind\":null,\"name\":\"background\"},{\"default\":3000,\"kind\":null,\"name\":\"duration\"},{\"default\":null,\"kind\":null,\"name\":\"icon\"},{\"default\":\"\",\"kind\":null,\"name\":\"message\"},{\"default\":null,\"kind\":null,\"name\":\"notification_type\"},{\"default\":false,\"kind\":null,\"name\":\"_destroyed\"}]},{\"extends\":null,\"module\":null,\"name\":\"TemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]},{\"extends\":null,\"module\":null,\"name\":\"MaterialTemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]}],\"roots\":{\"references\":[{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"2285\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"2264\",\"type\":\"Circle\"},{\"attributes\":{\"line_alpha\":0.6,\"line_color\":\"green\",\"line_width\":6,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2268\",\"type\":\"Line\"},{\"attributes\":{\"line_alpha\":0.2,\"line_color\":\"green\",\"line_width\":6,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2270\",\"type\":\"Line\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"2283\"},\"selection_policy\":{\"id\":\"2282\"}},\"id\":\"2267\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"source\":{\"id\":\"2267\"}},\"id\":\"2272\",\"type\":\"CDSView\"},{\"attributes\":{\"line_alpha\":0.1,\"line_color\":\"green\",\"line_width\":6,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2269\",\"type\":\"Line\"},{\"attributes\":{\"label\":{\"value\":\"True function\"},\"renderers\":[{\"id\":\"2271\"}]},\"id\":\"2286\",\"type\":\"LegendItem\"},{\"attributes\":{},\"id\":\"2256\",\"type\":\"UnionRenderers\"},{\"attributes\":{},\"id\":\"2282\",\"type\":\"UnionRenderers\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00124\",\"sizing_mode\":\"stretch_width\"},\"id\":\"2315\",\"type\":\"Spacer\"},{\"attributes\":{},\"id\":\"2257\",\"type\":\"Selection\"},{\"attributes\":{},\"id\":\"2251\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{},\"id\":\"2231\",\"type\":\"SaveTool\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"2220\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"2223\"},{\"id\":\"2227\"},{\"id\":\"2259\"}],\"height\":450,\"left\":[{\"id\":\"2224\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"2247\"},{\"id\":\"2271\"}],\"title\":{\"id\":\"2210\"},\"toolbar\":{\"id\":\"2235\"},\"toolbar_location\":null,\"x_range\":{\"id\":\"2212\"},\"x_scale\":{\"id\":\"2216\"},\"y_range\":{\"id\":\"2214\"},\"y_scale\":{\"id\":\"2218\"}},\"id\":\"2209\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{\"overlay\":{\"id\":\"2234\"}},\"id\":\"2230\",\"type\":\"BoxZoomTool\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"W79SoNavlD8yb2dIzNu5P2WTTTbZZMM/UetXCtT6xT88Q2LezpDIPxPzdobEvM0/ejtDYt7O0D910UUXXXTRP2ApUOtXCtQ/W79SoNav1D9WVVVVVVXVP1HrVwrU+tU/TIFav1Kg1j9GF1100UXXPzxDYt7OkNg/N9lkk0022T8yb2dIzNvZPy0Fav1Kgdo/HcdxHMdx3D8YXXTRRRfdPxPzdobEvN0/CB988MEH3z998MEHH3zgP3o7Q2LezuA/d4bEvJ0h4T910UUXXXThP3BnSMzbGeI/aEjM2xkS4z9gKVDrVwrkP1gK1PqVAuU/U6DWrxSo5T9ONtlkk03mP0Ri3s6QmOc/OY7jOI7j6D832WSTTTbpPzJvZ0jM2+k/L7rooosu6j8tBWr9SoHqPypQ61cK1Oo/J5tssskm6z8iMW9nSMzrPx3HcRzHcew/Faj1KwVq7T8T83aGxLztPxA++OCDD+4/Dol5O0Ni7j8L1PqVArXuPwZq/UqBWu8/A7V+pUCt7z8AAAAAAADwPw==\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[50]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[50]}},\"selected\":{\"id\":\"2257\"},\"selection_policy\":{\"id\":\"2256\"}},\"id\":\"2242\",\"type\":\"ColumnDataSource\"},{\"attributes\":{},\"id\":\"2228\",\"type\":\"PanTool\"},{\"attributes\":{},\"id\":\"2283\",\"type\":\"Selection\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"2260\"},{\"id\":\"2286\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"2259\",\"type\":\"Legend\"},{\"attributes\":{},\"id\":\"2229\",\"type\":\"WheelZoomTool\"},{\"attributes\":{\"source\":{\"id\":\"2242\"}},\"id\":\"2248\",\"type\":\"CDSView\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00123\",\"sizing_mode\":\"stretch_width\"},\"id\":\"2288\",\"type\":\"Spacer\"},{\"attributes\":{},\"id\":\"2212\",\"type\":\"DataRange1d\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2246\",\"type\":\"Circle\"},{\"attributes\":{\"axis\":{\"id\":\"2224\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"2227\",\"type\":\"Grid\"},{\"attributes\":{},\"id\":\"2254\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{},\"id\":\"2233\",\"type\":\"HelpTool\"},{\"attributes\":{\"tools\":[{\"id\":\"2228\"},{\"id\":\"2229\"},{\"id\":\"2230\"},{\"id\":\"2231\"},{\"id\":\"2232\"},{\"id\":\"2233\"},{\"id\":\"2261\"},{\"id\":\"2265\"}]},\"id\":\"2235\",\"type\":\"Toolbar\"},{\"attributes\":{\"axis\":{\"id\":\"2220\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"2223\",\"type\":\"Grid\"},{\"attributes\":{},\"id\":\"2218\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"2255\",\"type\":\"AllLabels\"},{\"attributes\":{\"client_comm_id\":\"cf04ffdb6fde46ffa28ef90c7e7342f5\",\"comm_id\":\"951ff1d9123041628e91a901a7414d70\",\"plot_id\":\"2287\"},\"id\":\"2355\",\"type\":\"panel.models.comm_manager.CommManager\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2243\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"2225\",\"type\":\"BasicTicker\"},{\"attributes\":{},\"id\":\"2252\",\"type\":\"AllLabels\"},{\"attributes\":{\"axis_label\":\"x\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"2254\"},\"group\":null,\"major_label_policy\":{\"id\":\"2255\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"2221\"}},\"id\":\"2220\",\"type\":\"LinearAxis\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"2247\"}]},\"id\":\"2260\",\"type\":\"LegendItem\"},{\"attributes\":{},\"id\":\"2232\",\"type\":\"ResetTool\"},{\"attributes\":{\"children\":[{\"id\":\"2288\"},{\"id\":\"2209\"},{\"id\":\"2315\"}],\"margin\":[0,0,0,0],\"name\":\"Row00127\"},\"id\":\"2287\",\"type\":\"Row\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Learning problem with groundtruth\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"2210\",\"type\":\"Title\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2245\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"2221\",\"type\":\"BasicTicker\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"2263\",\"type\":\"Circle\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"2285\"},\"renderers\":[{\"id\":\"2247\"}]},\"id\":\"2265\",\"type\":\"BoxSelectTool\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"2234\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"axis_label\":\"y\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"2251\"},\"group\":null,\"major_label_policy\":{\"id\":\"2252\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"2225\"}},\"id\":\"2224\",\"type\":\"LinearAxis\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"2242\"},\"glyph\":{\"id\":\"2243\"},\"group\":null,\"hover_glyph\":{\"id\":\"2245\"},\"muted_glyph\":{\"id\":\"2246\"},\"nonselection_glyph\":{\"id\":\"2264\"},\"selection_glyph\":{\"id\":\"2263\"},\"view\":{\"id\":\"2248\"}},\"id\":\"2247\",\"type\":\"GlyphRenderer\"},{\"attributes\":{},\"id\":\"2216\",\"type\":\"LinearScale\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"2247\"}],\"tooltips\":null},\"id\":\"2261\",\"type\":\"HoverTool\"},{\"attributes\":{},\"id\":\"2214\",\"type\":\"DataRange1d\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"2267\"},\"glyph\":{\"id\":\"2268\"},\"group\":null,\"hover_glyph\":null,\"muted_glyph\":{\"id\":\"2270\"},\"nonselection_glyph\":{\"id\":\"2269\"},\"view\":{\"id\":\"2272\"}},\"id\":\"2271\",\"type\":\"GlyphRenderer\"}],\"root_ids\":[\"2287\",\"2355\"]},\"title\":\"Bokeh Application\",\"version\":\"2.4.3\"}};\n",
       "    var render_items = [{\"docid\":\"d8321e7f-68be-4e59-83ca-6a297eb77df9\",\"root_ids\":[\"2287\"],\"roots\":{\"2287\":\"c0f2d06d-b2e2-4cc6-98d4-85472d4aa2da\"}}];\n",
       "    root.Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "    for (const render_item of render_items) {\n",
       "      for (const root_id of render_item.root_ids) {\n",
       "\tconst id_el = document.getElementById(root_id)\n",
       "\tif (id_el.children.length && (id_el.children[0].className === 'bk-root')) {\n",
       "\t  const root_el = id_el.children[0]\n",
       "\t  root_el.id = root_el.id + '-rendered'\n",
       "\t}\n",
       "      }\n",
       "    }\n",
       "  }\n",
       "  if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "        clearInterval(timer);\n",
       "        embed_document(root);\n",
       "      } else if (document.readyState == \"complete\") {\n",
       "        attempts++;\n",
       "        if (attempts > 200) {\n",
       "          clearInterval(timer);\n",
       "          console.log(\"Bokeh: ERROR: Unable to run BokehJS code because BokehJS library is missing\");\n",
       "        }\n",
       "      }\n",
       "    }, 25, root)\n",
       "  }\n",
       "})(window);</script>"
      ],
      "text/plain": [
       "Row\n",
       "    [0] HSpacer()\n",
       "    [1] Bokeh(Figure)\n",
       "    [2] HSpacer()"
      ]
     },
     "execution_count": 10,
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {
       "id": "2287"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Generating the data and subsampling\n",
    "x_all = np.linspace(0, 1, 100); x_plot = np.linspace(0, 1, 100)\n",
    "rng.shuffle(x_all); x = np.sort(x_all[:50])\n",
    "poly = np.poly1d([3,0,1])\n",
    "# Adding some external noise\n",
    "epsilon = np.random.uniform(-0.2,0.2,x.shape)\n",
    "y = poly(x)+ epsilon\n",
    "p = scatter(x, y, title=\"Learning problem with groundtruth\")\n",
    "p.line(x_plot, poly(x_plot), line_width=6, line_alpha=0.6, color=\"green\", legend_label=r\"True function\")\n",
    "plot = (center_plot(p))\n",
    "plot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summarizing our observations (interactive)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Creating the object\n"
     ]
    },
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.holoviews_exec.v0+json": "",
      "text/html": [
       "<div id='2479'>\n",
       "  <div class=\"bk-root\" id=\"aaaf79c9-ff83-4ac9-9f25-879b5185e5af\" data-root-id=\"2479\"></div>\n",
       "</div>\n",
       "<script type=\"application/javascript\">(function(root) {\n",
       "  function embed_document(root) {\n",
       "    var docs_json = {\"a6209349-f54c-439f-a1e9-11c22e1f881a\":{\"defs\":[{\"extends\":null,\"module\":null,\"name\":\"ReactiveHTML1\",\"overrides\":[],\"properties\":[]},{\"extends\":null,\"module\":null,\"name\":\"FlexBox1\",\"overrides\":[],\"properties\":[{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_content\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_items\"},{\"default\":\"row\",\"kind\":null,\"name\":\"flex_direction\"},{\"default\":\"wrap\",\"kind\":null,\"name\":\"flex_wrap\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"justify_content\"}]},{\"extends\":null,\"module\":null,\"name\":\"GridStack1\",\"overrides\":[],\"properties\":[{\"default\":\"warn\",\"kind\":null,\"name\":\"mode\"},{\"default\":null,\"kind\":null,\"name\":\"ncols\"},{\"default\":null,\"kind\":null,\"name\":\"nrows\"},{\"default\":true,\"kind\":null,\"name\":\"allow_resize\"},{\"default\":true,\"kind\":null,\"name\":\"allow_drag\"},{\"default\":[],\"kind\":null,\"name\":\"state\"}]},{\"extends\":null,\"module\":null,\"name\":\"click1\",\"overrides\":[],\"properties\":[{\"default\":\"\",\"kind\":null,\"name\":\"terminal_output\"},{\"default\":\"\",\"kind\":null,\"name\":\"debug_name\"},{\"default\":0,\"kind\":null,\"name\":\"clears\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationAreaBase1\",\"overrides\":[],\"properties\":[{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationArea1\",\"overrides\":[],\"properties\":[{\"default\":[],\"kind\":null,\"name\":\"notifications\"},{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"},{\"default\":[{\"background\":\"#ffc107\",\"icon\":{\"className\":\"fas fa-exclamation-triangle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"warning\"},{\"background\":\"#007bff\",\"icon\":{\"className\":\"fas fa-info-circle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"info\"}],\"kind\":null,\"name\":\"types\"}]},{\"extends\":null,\"module\":null,\"name\":\"Notification\",\"overrides\":[],\"properties\":[{\"default\":null,\"kind\":null,\"name\":\"background\"},{\"default\":3000,\"kind\":null,\"name\":\"duration\"},{\"default\":null,\"kind\":null,\"name\":\"icon\"},{\"default\":\"\",\"kind\":null,\"name\":\"message\"},{\"default\":null,\"kind\":null,\"name\":\"notification_type\"},{\"default\":false,\"kind\":null,\"name\":\"_destroyed\"}]},{\"extends\":null,\"module\":null,\"name\":\"TemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]},{\"extends\":null,\"module\":null,\"name\":\"MaterialTemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]}],\"roots\":{\"references\":[{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"2475\",\"type\":\"Circle\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"2454\"},\"glyph\":{\"id\":\"2455\"},\"group\":null,\"hover_glyph\":{\"id\":\"2457\"},\"muted_glyph\":{\"id\":\"2458\"},\"nonselection_glyph\":{\"id\":\"2476\"},\"selection_glyph\":{\"id\":\"2475\"},\"view\":{\"id\":\"2460\"}},\"id\":\"2459\",\"type\":\"GlyphRenderer\"},{\"attributes\":{\"args\":{\"bidirectional\":true,\"properties\":{\"readonly\":\"disabled\"},\"source\":{\"id\":\"2515\"},\"target\":{\"id\":\"2515\"}},\"code\":\"\\n    var value = source['disabled'];\\n    value = value;\\n    value = value;\\n    try {\\n      var property = target.properties['readonly'];\\n      if (property !== undefined) { property.validate(value); }\\n    } catch(err) {\\n      console.log('WARNING: Could not set readonly on target, raised error: ' + err);\\n      return;\\n    }\\n    try {\\n      target['readonly'] = value;\\n    } catch(err) {\\n      console.log(err)\\n    }\\n    \",\"tags\":[[140461258160592,[null,\"readonly\"],[null,\"disabled\"]]]},\"id\":\"2529\",\"type\":\"CustomJS\"},{\"attributes\":{},\"id\":\"2437\",\"type\":\"BasicTicker\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"2476\",\"type\":\"Circle\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"2446\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Polynomial regression problem\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"2422\",\"type\":\"Title\"},{\"attributes\":{\"tools\":[{\"id\":\"2440\"},{\"id\":\"2441\"},{\"id\":\"2442\"},{\"id\":\"2443\"},{\"id\":\"2444\"},{\"id\":\"2445\"},{\"id\":\"2473\"},{\"id\":\"2477\"}]},\"id\":\"2447\",\"type\":\"Toolbar\"},{\"attributes\":{},\"id\":\"2443\",\"type\":\"SaveTool\"},{\"attributes\":{\"axis_label\":\"$$\\\\color{white} -2.0x^2 + 0.0x^1 + 1.0x^0$$\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"2463\"},\"group\":null,\"major_label_policy\":{\"id\":\"2464\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"2437\"}},\"id\":\"2436\",\"type\":\"LinearAxis\"},{\"attributes\":{\"children\":[{\"id\":\"2480\"},{\"id\":\"2481\"},{\"id\":\"2516\"}],\"margin\":[0,0,0,0],\"name\":\"Row00162\"},\"id\":\"2479\",\"type\":\"Row\"},{\"attributes\":{},\"id\":\"2445\",\"type\":\"HelpTool\"},{\"attributes\":{},\"id\":\"2469\",\"type\":\"Selection\"},{\"attributes\":{\"end\":5.0,\"margin\":[5,10,5,10],\"start\":0.0,\"step\":0.1,\"title\":\"Noise level\",\"value\":0.1},\"id\":\"2486\",\"type\":\"Slider\"},{\"attributes\":{\"margin\":[5,10,5,10],\"name\":\"\",\"text\":\"<b>Problem parameters</b>\"},\"id\":\"2484\",\"type\":\"Div\"},{\"attributes\":{},\"id\":\"2441\",\"type\":\"WheelZoomTool\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2458\",\"type\":\"Circle\"},{\"attributes\":{\"args\":{\"bidirectional\":true,\"properties\":{\"readonly\":\"disabled\"},\"source\":{\"id\":\"2515\"},\"target\":{\"id\":\"2515\"}},\"code\":\"\\n    var value = source['readonly'];\\n    value = value;\\n    value = value;\\n    try {\\n      var property = target.properties['disabled'];\\n      if (property !== undefined) { property.validate(value); }\\n    } catch(err) {\\n      console.log('WARNING: Could not set disabled on target, raised error: ' + err);\\n      return;\\n    }\\n    try {\\n      target['disabled'] = value;\\n    } catch(err) {\\n      console.log(err)\\n    }\\n    \",\"tags\":[[140461258160592,[null,\"readonly\"],[null,\"disabled\"]]]},\"id\":\"2528\",\"type\":\"CustomJS\"},{\"attributes\":{\"axis_label\":\"$$\\\\color{white} x$$\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"2466\"},\"group\":null,\"major_label_policy\":{\"id\":\"2467\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"2433\"}},\"id\":\"2432\",\"type\":\"LinearAxis\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"2459\"}]},\"id\":\"2472\",\"type\":\"LegendItem\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"2503\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"end\":5.0,\"margin\":[5,10,5,10],\"start\":-5.0,\"step\":0.1,\"title\":\"X 1\",\"value\":0.0},\"id\":\"2490\",\"type\":\"Slider\"},{\"attributes\":{\"axis\":{\"id\":\"2436\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"2439\",\"type\":\"Grid\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2457\",\"type\":\"Circle\"},{\"attributes\":{\"overlay\":{\"id\":\"2446\"}},\"id\":\"2442\",\"type\":\"BoxZoomTool\"},{\"attributes\":{},\"id\":\"2433\",\"type\":\"BasicTicker\"},{\"attributes\":{},\"id\":\"2464\",\"type\":\"AllLabels\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2455\",\"type\":\"Circle\"},{\"attributes\":{\"children\":[{\"id\":\"2483\"},{\"id\":\"2492\"}],\"margin\":[0,0,0,0],\"name\":\"Row00157\"},\"id\":\"2482\",\"type\":\"Row\"},{\"attributes\":{\"children\":[{\"id\":\"2484\"},{\"id\":\"2485\"},{\"id\":\"2486\"},{\"id\":\"2487\"},{\"id\":\"2488\"},{\"id\":\"2489\"},{\"id\":\"2490\"},{\"id\":\"2491\"}],\"margin\":[5,5,5,5],\"min_width\":300,\"name\":\"Problem parameters\",\"width\":300},\"id\":\"2483\",\"type\":\"Column\"},{\"attributes\":{},\"id\":\"2466\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"code\":\"def polynomial(\\n        coefficients: jnp.ndarray,\\n        n_observations: int = 100) -> [jnp.ndarray, jnp.ndarray]:\\n    x = jnp.linspace(0, 1, n_observations)\\n    y = jnp.polyval(coefficients, x)\\n    return x, \",\"filename\":\"\",\"height\":150,\"js_property_callbacks\":{\"change:disabled\":[{\"id\":\"2529\"}],\"change:readonly\":[{\"id\":\"2528\"}]},\"language\":\"python\",\"margin\":[5,10,5,10],\"sizing_mode\":\"stretch_both\",\"theme\":\"chaos\"},\"id\":\"2515\",\"type\":\"panel.models.ace.AcePlot\"},{\"attributes\":{\"end\":5.0,\"margin\":[5,10,5,10],\"start\":-5.0,\"step\":0.1,\"title\":\"X 2\",\"value\":-2.0},\"id\":\"2489\",\"type\":\"Slider\"},{\"attributes\":{\"labels\":[\"Error\"],\"margin\":[5,10,5,10]},\"id\":\"2488\",\"type\":\"CheckboxGroup\"},{\"attributes\":{\"labels\":[\"Ground truth\"],\"margin\":[5,10,5,10]},\"id\":\"2487\",\"type\":\"CheckboxGroup\"},{\"attributes\":{},\"id\":\"2467\",\"type\":\"AllLabels\"},{\"attributes\":{\"end\":1000,\"margin\":[5,10,5,10],\"start\":10,\"title\":\"Number observations\",\"value\":100},\"id\":\"2485\",\"type\":\"Slider\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00131\",\"sizing_mode\":\"stretch_width\"},\"id\":\"2480\",\"type\":\"Spacer\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"2432\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"2435\"},{\"id\":\"2439\"},{\"id\":\"2471\"}],\"height\":450,\"left\":[{\"id\":\"2436\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"2459\"}],\"title\":{\"id\":\"2422\"},\"toolbar\":{\"id\":\"2447\"},\"toolbar_location\":null,\"x_range\":{\"id\":\"2424\"},\"x_scale\":{\"id\":\"2428\"},\"y_range\":{\"id\":\"2426\"},\"y_scale\":{\"id\":\"2430\"}},\"id\":\"2421\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{},\"id\":\"2430\",\"type\":\"LinearScale\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"2459\"}],\"tooltips\":null},\"id\":\"2473\",\"type\":\"HoverTool\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"AAAAALV+JTy1fqU8ED74PLV+JT1i3k49ED54Pd7OkD21fqU9jC66PWLezj05juM9ED74PfN2Bj7ezhA+yiYbPrV+JT6g1i8+jC46PneGRD5i3k4+TjZZPjmOYz4k5m0+ED54Pv1KgT7zdoY+6aKLPt7OkD7U+pU+yiabPr9SoD61fqU+q6qqPqDWrz6WArU+jC66PoFavz53hsQ+bbLJPmLezj5YCtQ+TjbZPkNi3j45juM+L7roPiTm7T4aEvM+ED74PgVq/T79SgE/+OADP/N2Bj/uDAk/6aILP+Q4Dj/ezhA/2WQTP9T6FT/PkBg/yiYbP8W8HT+/UiA/uugiP7V+JT+wFCg/q6oqP6VALT+g1i8/m2wyP5YCNT+RmDc/jC46P4bEPD+BWj8/fPBBP3eGRD9yHEc/bbJJP2dITD9i3k4/XXRRP1gKVD9ToFY/TjZZP0jMWz9DYl4/PvhgPzmOYz80JGY/L7poPylQaz8k5m0/H3xwPxoScz8VqHU/ED54PwrUej8Fan0/AACAPw==\",\"dtype\":\"float32\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float32\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"2469\"},\"selection_policy\":{\"id\":\"2468\"}},\"id\":\"2454\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"source\":{\"id\":\"2454\"}},\"id\":\"2460\",\"type\":\"CDSView\"},{\"attributes\":{},\"id\":\"2463\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{},\"id\":\"2444\",\"type\":\"ResetTool\"},{\"attributes\":{},\"id\":\"2468\",\"type\":\"UnionRenderers\"},{\"attributes\":{\"client_comm_id\":\"31f6dacb9f394365bc25e00a235c6211\",\"comm_id\":\"a46e3383437d4de8a84c09818351f8b8\",\"plot_id\":\"2479\"},\"id\":\"2552\",\"type\":\"panel.models.comm_manager.CommManager\"},{\"attributes\":{},\"id\":\"2424\",\"type\":\"DataRange1d\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"2472\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"2471\",\"type\":\"Legend\"},{\"attributes\":{},\"id\":\"2428\",\"type\":\"LinearScale\"},{\"attributes\":{\"children\":[{\"id\":\"2421\"}],\"margin\":[0,0,0,0],\"name\":\"Row00154\"},\"id\":\"2492\",\"type\":\"Row\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00161\",\"sizing_mode\":\"stretch_width\"},\"id\":\"2516\",\"type\":\"Spacer\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"2503\"},\"renderers\":[{\"id\":\"2459\"}]},\"id\":\"2477\",\"type\":\"BoxSelectTool\"},{\"attributes\":{\"end\":5.0,\"margin\":[5,10,5,10],\"start\":-5.0,\"step\":0.1,\"title\":\"X 0\",\"value\":1.0},\"id\":\"2491\",\"type\":\"Slider\"},{\"attributes\":{},\"id\":\"2426\",\"type\":\"DataRange1d\"},{\"attributes\":{\"axis\":{\"id\":\"2432\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"2435\",\"type\":\"Grid\"},{\"attributes\":{\"children\":[{\"id\":\"2482\"},{\"id\":\"2515\"}],\"margin\":[0,0,0,0],\"name\":\"Column00160\"},\"id\":\"2481\",\"type\":\"Column\"},{\"attributes\":{},\"id\":\"2440\",\"type\":\"PanTool\"}],\"root_ids\":[\"2479\",\"2552\"]},\"title\":\"Bokeh Application\",\"version\":\"2.4.3\"}};\n",
       "    var render_items = [{\"docid\":\"a6209349-f54c-439f-a1e9-11c22e1f881a\",\"root_ids\":[\"2479\"],\"roots\":{\"2479\":\"aaaf79c9-ff83-4ac9-9f25-879b5185e5af\"}}];\n",
       "    root.Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "    for (const render_item of render_items) {\n",
       "      for (const root_id of render_item.root_ids) {\n",
       "\tconst id_el = document.getElementById(root_id)\n",
       "\tif (id_el.children.length && (id_el.children[0].className === 'bk-root')) {\n",
       "\t  const root_el = id_el.children[0]\n",
       "\t  root_el.id = root_el.id + '-rendered'\n",
       "\t}\n",
       "      }\n",
       "    }\n",
       "  }\n",
       "  if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "        clearInterval(timer);\n",
       "        embed_document(root);\n",
       "      } else if (document.readyState == \"complete\") {\n",
       "        attempts++;\n",
       "        if (attempts > 200) {\n",
       "          clearInterval(timer);\n",
       "          console.log(\"Bokeh: ERROR: Unable to run BokehJS code because BokehJS library is missing\");\n",
       "        }\n",
       "      }\n",
       "    }, 25, root)\n",
       "  }\n",
       "})(window);</script>"
      ],
      "text/plain": [
       "Row\n",
       "    [0] HSpacer()\n",
       "    [1] Column\n",
       "        [0] Row\n",
       "            [0] Param(RegressionPolynomial, name='Problem parameters', widgets={})\n",
       "            [1] ParamMethod(method, _pane=Bokeh)\n",
       "        [1] Ace(height=150, language='python', sizing_mode='stretch_both', theme='chaos', value='def polynomial(\\n        ...)\n",
       "    [2] HSpacer()"
      ]
     },
     "execution_count": 11,
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {
       "id": "2479"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from cml.tasks import RegressionPolynomial\n",
    "explorer = RegressionPolynomial()\n",
    "explorer.render()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Using learning libraries (`scikit-learn`)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To get a first grip on what machine learning does, we will rely on the `scikit-learn` library. This contains already coded models and learning procedure, that will allow us to _learn_ the parameters of this unknown function.\n",
    "\n",
    "Here we already know that we want to use a `PolynomialFeatures` model to perfom `LinearRegression` and that this polynomial should be of degree 2."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "from sklearn.pipeline import make_pipeline\n",
    "# Our data to fit\n",
    "X = x[:, np.newaxis]\n",
    "# Degree of our polynomial\n",
    "degree = 30;\n",
    "# Create our polynomial model for regression\n",
    "model = make_pipeline(PolynomialFeatures(degree), LinearRegression())\n",
    "# Fit the parameters of this model\n",
    "model.fit(X, y);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have trained the model, we can perform _predictions_ from it, meaning that we can infer the output of the function at values that we did not observe originally."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model error : 0.08910089822161481\n"
     ]
    },
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.holoviews_exec.v0+json": "",
      "text/html": [
       "<div id='2686'>\n",
       "  <div class=\"bk-root\" id=\"636dc7dc-8d6a-4362-9585-c7421a5aa039\" data-root-id=\"2686\"></div>\n",
       "</div>\n",
       "<script type=\"application/javascript\">(function(root) {\n",
       "  function embed_document(root) {\n",
       "    var docs_json = {\"014b63b3-16d1-4413-94b5-837ee28e67b8\":{\"defs\":[{\"extends\":null,\"module\":null,\"name\":\"ReactiveHTML1\",\"overrides\":[],\"properties\":[]},{\"extends\":null,\"module\":null,\"name\":\"FlexBox1\",\"overrides\":[],\"properties\":[{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_content\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_items\"},{\"default\":\"row\",\"kind\":null,\"name\":\"flex_direction\"},{\"default\":\"wrap\",\"kind\":null,\"name\":\"flex_wrap\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"justify_content\"}]},{\"extends\":null,\"module\":null,\"name\":\"GridStack1\",\"overrides\":[],\"properties\":[{\"default\":\"warn\",\"kind\":null,\"name\":\"mode\"},{\"default\":null,\"kind\":null,\"name\":\"ncols\"},{\"default\":null,\"kind\":null,\"name\":\"nrows\"},{\"default\":true,\"kind\":null,\"name\":\"allow_resize\"},{\"default\":true,\"kind\":null,\"name\":\"allow_drag\"},{\"default\":[],\"kind\":null,\"name\":\"state\"}]},{\"extends\":null,\"module\":null,\"name\":\"click1\",\"overrides\":[],\"properties\":[{\"default\":\"\",\"kind\":null,\"name\":\"terminal_output\"},{\"default\":\"\",\"kind\":null,\"name\":\"debug_name\"},{\"default\":0,\"kind\":null,\"name\":\"clears\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationAreaBase1\",\"overrides\":[],\"properties\":[{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationArea1\",\"overrides\":[],\"properties\":[{\"default\":[],\"kind\":null,\"name\":\"notifications\"},{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"},{\"default\":[{\"background\":\"#ffc107\",\"icon\":{\"className\":\"fas fa-exclamation-triangle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"warning\"},{\"background\":\"#007bff\",\"icon\":{\"className\":\"fas fa-info-circle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"info\"}],\"kind\":null,\"name\":\"types\"}]},{\"extends\":null,\"module\":null,\"name\":\"Notification\",\"overrides\":[],\"properties\":[{\"default\":null,\"kind\":null,\"name\":\"background\"},{\"default\":3000,\"kind\":null,\"name\":\"duration\"},{\"default\":null,\"kind\":null,\"name\":\"icon\"},{\"default\":\"\",\"kind\":null,\"name\":\"message\"},{\"default\":null,\"kind\":null,\"name\":\"notification_type\"},{\"default\":false,\"kind\":null,\"name\":\"_destroyed\"}]},{\"extends\":null,\"module\":null,\"name\":\"TemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]},{\"extends\":null,\"module\":null,\"name\":\"MaterialTemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]}],\"roots\":{\"references\":[{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2645\",\"type\":\"Circle\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"2682\"},\"selection_policy\":{\"id\":\"2681\"}},\"id\":\"2666\",\"type\":\"ColumnDataSource\"},{\"attributes\":{},\"id\":\"2611\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"2630\",\"type\":\"SaveTool\"},{\"attributes\":{\"overlay\":{\"id\":\"2633\"}},\"id\":\"2629\",\"type\":\"BoxZoomTool\"},{\"attributes\":{},\"id\":\"2682\",\"type\":\"Selection\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2642\",\"type\":\"Circle\"},{\"attributes\":{\"line_alpha\":0.6,\"line_color\":\"red\",\"line_width\":6,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2667\",\"type\":\"Line\"},{\"attributes\":{},\"id\":\"2627\",\"type\":\"PanTool\"},{\"attributes\":{},\"id\":\"2628\",\"type\":\"WheelZoomTool\"},{\"attributes\":{},\"id\":\"2624\",\"type\":\"BasicTicker\"},{\"attributes\":{},\"id\":\"2655\",\"type\":\"UnionRenderers\"},{\"attributes\":{\"axis\":{\"id\":\"2623\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"2626\",\"type\":\"Grid\"},{\"attributes\":{},\"id\":\"2654\",\"type\":\"AllLabels\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2644\",\"type\":\"Circle\"},{\"attributes\":{\"children\":[{\"id\":\"2687\"},{\"id\":\"2608\"},{\"id\":\"2714\"}],\"margin\":[0,0,0,0],\"name\":\"Row00169\"},\"id\":\"2686\",\"type\":\"Row\"},{\"attributes\":{},\"id\":\"2632\",\"type\":\"HelpTool\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00166\",\"sizing_mode\":\"stretch_width\"},\"id\":\"2714\",\"type\":\"Spacer\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00165\",\"sizing_mode\":\"stretch_width\"},\"id\":\"2687\",\"type\":\"Spacer\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"2662\",\"type\":\"Circle\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"2684\"},\"renderers\":[{\"id\":\"2646\"}]},\"id\":\"2664\",\"type\":\"BoxSelectTool\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"2646\"}],\"tooltips\":null},\"id\":\"2660\",\"type\":\"HoverTool\"},{\"attributes\":{\"source\":{\"id\":\"2666\"}},\"id\":\"2671\",\"type\":\"CDSView\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"2633\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"client_comm_id\":\"58c52d0a3b624ff8a238801e45fb27b8\",\"comm_id\":\"2bf103890f0e46c6bab4696634305799\",\"plot_id\":\"2686\"},\"id\":\"2754\",\"type\":\"panel.models.comm_manager.CommManager\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"2646\"}]},\"id\":\"2659\",\"type\":\"LegendItem\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"2663\",\"type\":\"Circle\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"2641\"},\"glyph\":{\"id\":\"2642\"},\"group\":null,\"hover_glyph\":{\"id\":\"2644\"},\"muted_glyph\":{\"id\":\"2645\"},\"nonselection_glyph\":{\"id\":\"2663\"},\"selection_glyph\":{\"id\":\"2662\"},\"view\":{\"id\":\"2647\"}},\"id\":\"2646\",\"type\":\"GlyphRenderer\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"2684\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"2666\"},\"glyph\":{\"id\":\"2667\"},\"group\":null,\"hover_glyph\":null,\"muted_glyph\":{\"id\":\"2669\"},\"nonselection_glyph\":{\"id\":\"2668\"},\"view\":{\"id\":\"2671\"}},\"id\":\"2670\",\"type\":\"GlyphRenderer\"},{\"attributes\":{},\"id\":\"2651\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"2615\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"2656\",\"type\":\"Selection\"},{\"attributes\":{\"line_alpha\":0.2,\"line_color\":\"red\",\"line_width\":6,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2669\",\"type\":\"Line\"},{\"attributes\":{},\"id\":\"2653\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"line_alpha\":0.1,\"line_color\":\"red\",\"line_width\":6,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2668\",\"type\":\"Line\"},{\"attributes\":{\"label\":{\"value\":\"Trained model\"},\"renderers\":[{\"id\":\"2670\"}]},\"id\":\"2685\",\"type\":\"LegendItem\"},{\"attributes\":{\"axis_label\":\"x\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"2653\"},\"group\":null,\"major_label_policy\":{\"id\":\"2654\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"2620\"}},\"id\":\"2619\",\"type\":\"LinearAxis\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Training a scikit-learn model\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"2609\",\"type\":\"Title\"},{\"attributes\":{},\"id\":\"2620\",\"type\":\"BasicTicker\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[50]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[50]}},\"selected\":{\"id\":\"2656\"},\"selection_policy\":{\"id\":\"2655\"}},\"id\":\"2641\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"axis_label\":\"y\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"2650\"},\"group\":null,\"major_label_policy\":{\"id\":\"2651\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"2624\"}},\"id\":\"2623\",\"type\":\"LinearAxis\"},{\"attributes\":{},\"id\":\"2631\",\"type\":\"ResetTool\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"2619\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"2622\"},{\"id\":\"2626\"},{\"id\":\"2658\"}],\"height\":450,\"left\":[{\"id\":\"2623\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"2646\"},{\"id\":\"2670\"}],\"title\":{\"id\":\"2609\"},\"toolbar\":{\"id\":\"2634\"},\"toolbar_location\":null,\"x_range\":{\"id\":\"2611\"},\"x_scale\":{\"id\":\"2615\"},\"y_range\":{\"id\":\"2613\"},\"y_scale\":{\"id\":\"2617\"}},\"id\":\"2608\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{\"source\":{\"id\":\"2641\"}},\"id\":\"2647\",\"type\":\"CDSView\"},{\"attributes\":{},\"id\":\"2681\",\"type\":\"UnionRenderers\"},{\"attributes\":{},\"id\":\"2617\",\"type\":\"LinearScale\"},{\"attributes\":{\"tools\":[{\"id\":\"2627\"},{\"id\":\"2628\"},{\"id\":\"2629\"},{\"id\":\"2630\"},{\"id\":\"2631\"},{\"id\":\"2632\"},{\"id\":\"2660\"},{\"id\":\"2664\"}]},\"id\":\"2634\",\"type\":\"Toolbar\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"2659\"},{\"id\":\"2685\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"2658\",\"type\":\"Legend\"},{\"attributes\":{\"axis\":{\"id\":\"2619\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"2622\",\"type\":\"Grid\"},{\"attributes\":{},\"id\":\"2650\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{},\"id\":\"2613\",\"type\":\"DataRange1d\"}],\"root_ids\":[\"2686\",\"2754\"]},\"title\":\"Bokeh Application\",\"version\":\"2.4.3\"}};\n",
       "    var render_items = [{\"docid\":\"014b63b3-16d1-4413-94b5-837ee28e67b8\",\"root_ids\":[\"2686\"],\"roots\":{\"2686\":\"636dc7dc-8d6a-4362-9585-c7421a5aa039\"}}];\n",
       "    root.Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "    for (const render_item of render_items) {\n",
       "      for (const root_id of render_item.root_ids) {\n",
       "\tconst id_el = document.getElementById(root_id)\n",
       "\tif (id_el.children.length && (id_el.children[0].className === 'bk-root')) {\n",
       "\t  const root_el = id_el.children[0]\n",
       "\t  root_el.id = root_el.id + '-rendered'\n",
       "\t}\n",
       "      }\n",
       "    }\n",
       "  }\n",
       "  if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "        clearInterval(timer);\n",
       "        embed_document(root);\n",
       "      } else if (document.readyState == \"complete\") {\n",
       "        attempts++;\n",
       "        if (attempts > 200) {\n",
       "          clearInterval(timer);\n",
       "          console.log(\"Bokeh: ERROR: Unable to run BokehJS code because BokehJS library is missing\");\n",
       "        }\n",
       "      }\n",
       "    }, 25, root)\n",
       "  }\n",
       "})(window);</script>"
      ],
      "text/plain": [
       "Row\n",
       "    [0] HSpacer()\n",
       "    [1] Bokeh(Figure)\n",
       "    [2] HSpacer()"
      ]
     },
     "execution_count": 13,
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {
       "id": "2686"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Inference points (not observed)\n",
    "X_plot = x_plot[:, np.newaxis]\n",
    "# Predict the values\n",
    "y_plot = model.predict(X_plot)\n",
    "# Compute the error of our model at observed points\n",
    "Y_model_err = np.sqrt(np.mean(np.square(y-model.predict(X))))\n",
    "print(f'Model error : {Y_model_err}')\n",
    "# Plot the result\n",
    "p = scatter(x, y, title=\"Training a scikit-learn model\")\n",
    "p.line(x_plot, poly(x_plot), line_width=6, line_alpha=0.6, color=\"red\", legend_label=r\"Trained model\")\n",
    "plot = (center_plot(p))\n",
    "plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "from cml.tasks import RegressionPolynomialSolver\n",
    "explorer = RegressionPolynomialSolver()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.holoviews_exec.v0+json": "",
      "text/html": [
       "<div id='2884'>\n",
       "  <div class=\"bk-root\" id=\"b4045240-c9e1-44b9-86c6-e4ca34854d40\" data-root-id=\"2884\"></div>\n",
       "</div>\n",
       "<script type=\"application/javascript\">(function(root) {\n",
       "  function embed_document(root) {\n",
       "    var docs_json = {\"029f4470-e447-48bc-b39f-9c1cfce89c7e\":{\"defs\":[{\"extends\":null,\"module\":null,\"name\":\"ReactiveHTML1\",\"overrides\":[],\"properties\":[]},{\"extends\":null,\"module\":null,\"name\":\"FlexBox1\",\"overrides\":[],\"properties\":[{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_content\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_items\"},{\"default\":\"row\",\"kind\":null,\"name\":\"flex_direction\"},{\"default\":\"wrap\",\"kind\":null,\"name\":\"flex_wrap\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"justify_content\"}]},{\"extends\":null,\"module\":null,\"name\":\"GridStack1\",\"overrides\":[],\"properties\":[{\"default\":\"warn\",\"kind\":null,\"name\":\"mode\"},{\"default\":null,\"kind\":null,\"name\":\"ncols\"},{\"default\":null,\"kind\":null,\"name\":\"nrows\"},{\"default\":true,\"kind\":null,\"name\":\"allow_resize\"},{\"default\":true,\"kind\":null,\"name\":\"allow_drag\"},{\"default\":[],\"kind\":null,\"name\":\"state\"}]},{\"extends\":null,\"module\":null,\"name\":\"click1\",\"overrides\":[],\"properties\":[{\"default\":\"\",\"kind\":null,\"name\":\"terminal_output\"},{\"default\":\"\",\"kind\":null,\"name\":\"debug_name\"},{\"default\":0,\"kind\":null,\"name\":\"clears\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationAreaBase1\",\"overrides\":[],\"properties\":[{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationArea1\",\"overrides\":[],\"properties\":[{\"default\":[],\"kind\":null,\"name\":\"notifications\"},{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"},{\"default\":[{\"background\":\"#ffc107\",\"icon\":{\"className\":\"fas fa-exclamation-triangle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"warning\"},{\"background\":\"#007bff\",\"icon\":{\"className\":\"fas fa-info-circle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"info\"}],\"kind\":null,\"name\":\"types\"}]},{\"extends\":null,\"module\":null,\"name\":\"Notification\",\"overrides\":[],\"properties\":[{\"default\":null,\"kind\":null,\"name\":\"background\"},{\"default\":3000,\"kind\":null,\"name\":\"duration\"},{\"default\":null,\"kind\":null,\"name\":\"icon\"},{\"default\":\"\",\"kind\":null,\"name\":\"message\"},{\"default\":null,\"kind\":null,\"name\":\"notification_type\"},{\"default\":false,\"kind\":null,\"name\":\"_destroyed\"}]},{\"extends\":null,\"module\":null,\"name\":\"TemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]},{\"extends\":null,\"module\":null,\"name\":\"MaterialTemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]}],\"roots\":{\"references\":[{\"attributes\":{},\"id\":\"2863\",\"type\":\"AllLabels\"},{\"attributes\":{\"client_comm_id\":\"49a5712620014e2c89e205193738b954\",\"comm_id\":\"68843fd4df054e0a912f322bc4388650\",\"plot_id\":\"2884\"},\"id\":\"2968\",\"type\":\"panel.models.comm_manager.CommManager\"},{\"attributes\":{},\"id\":\"2844\",\"type\":\"HelpTool\"},{\"attributes\":{\"axis_label\":\"$$\\\\color{white} -2.0x^2 + 0.0x^1 + 1.0x^0$$\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"2862\"},\"group\":null,\"major_label_policy\":{\"id\":\"2863\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"2836\"}},\"id\":\"2835\",\"type\":\"LinearAxis\"},{\"attributes\":{},\"id\":\"2865\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"axis\":{\"id\":\"2831\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"2834\",\"type\":\"Grid\"},{\"attributes\":{\"axis\":{\"id\":\"2835\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"2838\",\"type\":\"Grid\"},{\"attributes\":{},\"id\":\"2866\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"2836\",\"type\":\"BasicTicker\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2857\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"2840\",\"type\":\"WheelZoomTool\"},{\"attributes\":{},\"id\":\"2839\",\"type\":\"PanTool\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"AAAAAAAAAABn0LLjOZV0P2fQsuM5lYQ/mjiM1dbfjj9n0LLjOZWUP4GEn1yIupk/mjiM1dbfnj9adjynkgKiP2fQsuM5laQ/dCopIOEnpz+BhJ9ciLqpP47eFZkvTaw/mjiM1dbfrj9USQEJP7mwP1p2PKeSArI/YaN3ReZLsz9n0LLjOZW0P2397YGN3rU/dCopIOEntz96V2S+NHG4P4GEn1yIurk/h7Ha+tsDuz+O3hWZL028P5QLUTeDlr0/mjiM1dbfvj/QsuM5lRTAP1RJAQk/ucA/198e2OhdwT9adjynkgLCP90MWnY8p8I/YaN3ReZLwz/kOZUUkPDDP2fQsuM5lcQ/6mbQsuM5xT9t/e2Bjd7FP/GTC1E3g8Y/dCopIOEnxz/3wEbviszHP3pXZL40ccg//u2Bjd4VyT+BhJ9ciLrJPwQbvSsyX8o/h7Ha+tsDyz8KSPjJhajLP47eFZkvTcw/EXUzaNnxzD+UC1E3g5bNPxeibgYtO84/mjiM1dbfzj8ez6mkgITPP9Cy4zmVFNA/En5yIepm0D9USQEJP7nQP5UUkPCTC9E/198e2Ohd0T8Zq62/PbDRP1p2PKeSAtI/nEHLjudU0j/dDFp2PKfSPx/Y6F2R+dI/YaN3ReZL0z+ibgYtO57TP+Q5lRSQ8NM/JQUk/ORC1D9n0LLjOZXUP6mbQcuO59Q/6mbQsuM51T8sMl+aOIzVP2397YGN3tU/r8h8aeIw1j/xkwtRN4PWPzJfmjiM1dY/dCopIOEn1z+19bcHNnrXP/fARu+KzNc/OYzV1t8e2D96V2S+NHHYP7wi86WJw9g//u2Bjd4V2T8/uRB1M2jZP4GEn1yIutk/wk8uRN0M2j8EG70rMl/aP0bmSxOHsdo/h7Ha+tsD2z/JfGniMFbbPwpI+MmFqNs/TBOHsdr62z+O3hWZL03cP8+ppICEn9w/EXUzaNnx3D9SQMJPLkTdP5QLUTeDlt0/1tbfHtjo3T8Xom4GLTveP1lt/e2Bjd4/mjiM1dbf3j/cAxu9KzLfPx7PqaSAhN8/X5o4jNXW3z/QsuM5lRTgP3EYq62/PeA/En5yIepm4D+z4zmVFJDgP1RJAQk/ueA/9K7IfGni4D+VFJDwkwvhPzZ6V2S+NOE/198e2Ohd4T94ReZLE4fhPxmrrb89sOE/uRB1M2jZ4T9adjynkgLiP/vbAxu9K+I/nEHLjudU4j89p5ICEn7iP90MWnY8p+I/fnIh6mbQ4j8f2OhdkfniP8A9sNG7IuM/YaN3ReZL4z8BCT+5EHXjP6JuBi07nuM/Q9TNoGXH4z/kOZUUkPDjP4WfXIi6GeQ/JQUk/ORC5D/GautvD2zkP2fQsuM5leQ/CDZ6V2S+5D+pm0HLjufkP0kBCT+5EOU/6mbQsuM55T+LzJcmDmPlPywyX5o4jOU/zZcmDmO15T9t/e2Bjd7lPw5jtfW3B+Y/r8h8aeIw5j9QLkTdDFrmP/GTC1E3g+Y/kfnSxGGs5j8yX5o4jNXmP9PEYay2/uY/dCopIOEn5z8VkPCTC1HnP7X1twc2euc/Vlt/e2Cj5z/3wEbvisznP5gmDmO19ec/OYzV1t8e6D/a8ZxKCkjoP3pXZL40ceg/G70rMl+a6D+8IvOlicPoP12Iuhm07Og//u2Bjd4V6T+eU0kBCT/pPz+5EHUzaOk/4B7Y6F2R6T+BhJ9ciLrpPyLqZtCy4+k/wk8uRN0M6j9jtfW3BzbqPwQbvSsyX+o/pYCEn1yI6j9G5ksTh7HqP+ZLE4ex2uo/h7Ha+tsD6z8oF6JuBi3rP8l8aeIwVus/auIwVlt/6z8KSPjJhajrP6utvz2w0es/TBOHsdr66z/teE4lBSTsP47eFZkvTew/LkTdDFp27D/PqaSAhJ/sP3APbPSuyOw/EXUzaNnx7D+y2vrbAxvtP1JAwk8uRO0/86WJw1ht7T+UC1E3g5btPzVxGKutv+0/1tbfHtjo7T92PKeSAhLuPxeibgYtO+4/uAc2eldk7j9Zbf3tgY3uP/rSxGGstu4/mjiM1dbf7j87nlNJAQnvP9wDG70rMu8/fWniMFZb7z8ez6mkgITvP780cRirre8/X5o4jNXW7z8AAAAAAADwPw==\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[200]},\"y\":{\"__ndarray__\":\"AAAAwDVF9T/dDFp2BBz1P7sZtCzT8vQ/mCYO46HJ9D91M2iZcKD0P1JAwk8/d/Q/ME0cBg5O9D8NWna83CT0P+pm0HKr+/M/x3MqKXrS8z+lgITfSKnzP4KN3pUXgPM/X5o4TOZW8z88p5ICtS3zPxq07LiDBPM/98BGb1Lb8j/UzaAlIbLyP7La+tvviPI/j+dUkr5f8j9s9K5IjTbyP0kBCf9bDfI/Jw5jtSrk8T8EG71r+brxP+EnFyLIkfE/vjRx2JZo8T+cQcuOZT/xP3lOJUU0FvE/Vlt/+wLt8D80aNmx0cPwPxF1M2igmvA/7oGNHm9x8D/LjufUPUjwP6ibQYsMH/A/DFE3g7br7z/GauvvU5nvP4CEn1zxRu8/O55TyY707j/2twc2LKLuP7DRu6LJT+4/autvD2f97T8lBSR8BKvtP+Ae2OihWO0/mjiMVT8G7T9VUkDC3LPsPw9s9C56Yew/yoWomxcP7D+En1wItbzrPz+5EHVSaus/+tLE4e8X6z+07HhOjcXqP28GLbsqc+o/KSDhJ8gg6j/jOZWUZc7pP55TSQEDfOk/WW39baAp6T8Th7HaPdfoP86gZUfbhOg/iLoZtHgy6D9D1M0gFuDnP/3tgY2zjec/twc2+lA75z9zIepm7ujmPy07ntOLluY/6FRSQClE5j+ibgatxvHlP1yIuhlkn+U/F6JuhgFN5T/RuyLznvrkP43V1l88qOQ/R++KzNlV5D8BCT85dwPkP7wi86UUseM/djynErJe4z8xVlt/TwzjP+tvD+zsueI/ponDWIpn4j9ho3fFJxXiPxu9KzLFwuE/1dbfnmJw4T+Q8JMLAB7hP0oKSHidy+A/BST85Dp54D/APbBR2CbgP/SuyHzrqN8/auIwViYE3z/eFZkvYV/eP1RJAQmcut0/yHxp4tYV3T88sNG7EXHcP7TjOZVMzNs/KBeibocn2z+eSgpIwoLaPxJ+ciH93dk/hrHa+jc52T/85ELUcpTYP3AYq62t79c/6EsTh+hK1z9cf3tgI6bWP9Cy4zleAdY/RuZLE5lc1T+8GbTs07fUPzBNHMYOE9Q/pICEn0lu0z8YtOx4hMnSP4znVFK/JNI/BBu9K/p/0T94TiUFNdvQP/CBjd5vNtA/yGrrb1Ujzz+w0bsiy9nNP5g4jNVAkMw/iJ9ciLZGyz9wBi07LP3JP1ht/e2hs8g/QNTNoBdqxz8oO55TjSDGPxiibgYD18Q/AAk/uXiNwz/obw9s7kPCP9DW3x5k+sA/cHtgo7Nhvz9gSQEJn868PzAXom6KO7o/AOVC1HWotz/QsuM5YRW1P6CAhJ9MgrI/AJ1KCnDerz+gOIzVRriqP0DUzaAdkqU/4G8PbPRroD8AF6JulouWPwCdSgqIfog/AFyIuhkvXz8AhqibwbKAv8ALUTezpZK/gNTNoAXynL9gTiUFLJ+jv8Cy4zlVxai/IBeibn7rrb/APbDR04ixv/BvD2zoG7S/EKJuBv2utr9A1M2gEUK5v3AGLTsm1bu/oDiM1Tpovr9otfW3p33Av3hOJQUyx8G/kOdUUrwQw7+ogISfRlrEv8AZtOzQo8W/0LLjOVvtxr/oSxOH5TbIv/jkQtRvgMm/EH5yIfrJyr8oF6JuhBPMv0Cw0bsOXc2/WEkBCZmmzr9o4jBWI/DPv8A9sNHWnNC/TApI+JtB0b/Y1t8eYebRv2Sjd0Umi9K/7G8PbOsv0794PKeSsNTTvwAJP7l1edS/jNXW3zoe1b8Yom4GAMPVv6BuBi3FZ9a/LDueU4oM17+4BzZ6T7HXv0TUzaAUVti/0KBlx9n62L9Ybf3tnp/Zv+Q5lRRkRNq/cAYtOynp2r/80sRh7o3bv4ifXIizMty/EGz0rnjX3L+YOIzVPXzdvyQFJPwCId6/sNG7IsjF3r88nlNJjWrfv2K19TepB+C/qJtBywta4L/ugY1ebqzgvzRo2fHQ/uC/ek4lhTNR4b++NHEYlqPhvwQbvav49eG/SgEJP1tI4r+Q51TSvZriv9bNoGUg7eK/GLTs+II/479emjiM5ZHjv6SAhB9I5OO/6mbQsqo25L8wTRxGDYnkv3YzaNlv2+S/uhm0bNIt5b8AAAAANYDlvw==\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[200]}},\"selected\":{\"id\":\"2909\"},\"selection_policy\":{\"id\":\"2908\"}},\"id\":\"2878\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"overlay\":{\"id\":\"2845\"}},\"id\":\"2841\",\"type\":\"BoxZoomTool\"},{\"attributes\":{},\"id\":\"2842\",\"type\":\"SaveTool\"},{\"attributes\":{},\"id\":\"2908\",\"type\":\"UnionRenderers\"},{\"attributes\":{},\"id\":\"2843\",\"type\":\"ResetTool\"},{\"attributes\":{},\"id\":\"2867\",\"type\":\"UnionRenderers\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"2845\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"2853\"},\"glyph\":{\"id\":\"2854\"},\"group\":null,\"hover_glyph\":{\"id\":\"2856\"},\"muted_glyph\":{\"id\":\"2857\"},\"nonselection_glyph\":{\"id\":\"2875\"},\"selection_glyph\":{\"id\":\"2874\"},\"view\":{\"id\":\"2859\"}},\"id\":\"2858\",\"type\":\"GlyphRenderer\"},{\"attributes\":{},\"id\":\"2868\",\"type\":\"Selection\"},{\"attributes\":{},\"id\":\"2909\",\"type\":\"Selection\"},{\"attributes\":{},\"id\":\"2829\",\"type\":\"LinearScale\"},{\"attributes\":{\"source\":{\"id\":\"2878\"}},\"id\":\"2883\",\"type\":\"CDSView\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Polynomial regression problem\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"2821\",\"type\":\"Title\"},{\"attributes\":{\"args\":{\"bidirectional\":true,\"properties\":{\"readonly\":\"disabled\"},\"source\":{\"id\":\"2925\"},\"target\":{\"id\":\"2925\"}},\"code\":\"\\n    var value = source['readonly'];\\n    value = value;\\n    value = value;\\n    try {\\n      var property = target.properties['disabled'];\\n      if (property !== undefined) { property.validate(value); }\\n    } catch(err) {\\n      console.log('WARNING: Could not set disabled on target, raised error: ' + err);\\n      return;\\n    }\\n    try {\\n      target['disabled'] = value;\\n    } catch(err) {\\n      console.log(err)\\n    }\\n    \",\"tags\":[[140461255846400,[null,\"readonly\"],[null,\"disabled\"]]]},\"id\":\"2940\",\"type\":\"CustomJS\"},{\"attributes\":{\"tools\":[{\"id\":\"2839\"},{\"id\":\"2840\"},{\"id\":\"2841\"},{\"id\":\"2842\"},{\"id\":\"2843\"},{\"id\":\"2844\"},{\"id\":\"2872\"},{\"id\":\"2876\"}]},\"id\":\"2846\",\"type\":\"Toolbar\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2854\",\"type\":\"Circle\"},{\"attributes\":{\"args\":{\"bidirectional\":true,\"properties\":{\"readonly\":\"disabled\"},\"source\":{\"id\":\"2925\"},\"target\":{\"id\":\"2925\"}},\"code\":\"\\n    var value = source['disabled'];\\n    value = value;\\n    value = value;\\n    try {\\n      var property = target.properties['readonly'];\\n      if (property !== undefined) { property.validate(value); }\\n    } catch(err) {\\n      console.log('WARNING: Could not set readonly on target, raised error: ' + err);\\n      return;\\n    }\\n    try {\\n      target['readonly'] = value;\\n    } catch(err) {\\n      console.log(err)\\n    }\\n    \",\"tags\":[[140461255846400,[null,\"readonly\"],[null,\"disabled\"]]]},\"id\":\"2941\",\"type\":\"CustomJS\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2856\",\"type\":\"Circle\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"2911\",\"type\":\"BoxAnnotation\"},{\"attributes\":{},\"id\":\"2823\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"2827\",\"type\":\"LinearScale\"},{\"attributes\":{\"end\":5.0,\"margin\":[5,10,5,10],\"start\":-5.0,\"step\":0.1,\"title\":\"X 1\",\"value\":0.0},\"id\":\"2895\",\"type\":\"Slider\"},{\"attributes\":{},\"id\":\"2825\",\"type\":\"DataRange1d\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"2831\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"2834\"},{\"id\":\"2838\"},{\"id\":\"2870\"}],\"height\":450,\"left\":[{\"id\":\"2835\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"2858\"},{\"id\":\"2882\"}],\"title\":{\"id\":\"2821\"},\"toolbar\":{\"id\":\"2846\"},\"toolbar_location\":null,\"x_range\":{\"id\":\"2823\"},\"x_scale\":{\"id\":\"2827\"},\"y_range\":{\"id\":\"2825\"},\"y_scale\":{\"id\":\"2829\"}},\"id\":\"2820\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{\"children\":[{\"id\":\"2885\"},{\"id\":\"2886\"},{\"id\":\"2926\"}],\"margin\":[0,0,0,0],\"name\":\"Row00206\"},\"id\":\"2884\",\"type\":\"Row\"},{\"attributes\":{\"line_alpha\":0.7,\"line_color\":\"red\",\"line_width\":4,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2879\",\"type\":\"Line\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00205\",\"sizing_mode\":\"stretch_width\"},\"id\":\"2926\",\"type\":\"Spacer\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float32\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"IDeMP+jsbz+WJYM/l/uKP8uxfD8FqmU/PvZrP7IOgD8tbYY/x62CP2aJdT9BXoA/bD5kP4LkXj94Vns/cRZkP5Zjgj94L2w/oMl2P8LMcT9h1GM/rzt+Pwojdj8q3Gg/Lk9gPyZ6VT8ewVQ/3QpvP8PtRT8ANkw/lPFhPyK3ZT9bOEc/oQ83P3bXOT8nVjY/mm5HP60+Ij9kwEs/KNcrP1O0Kj84JRE/p8MdP54yEj8i2TE/Wd0MP7eLBz/fyxg/rf4XPxtK2j7sLuU+WWfCPtzztz7ISeA+TIP3PoRXuj7Osao+EkKcPguepD7g5ZQ+RuSKPk7ZNT5ElWI+rdgoPuQbaD6RKR0+wUcyPiLLGz765ws+fab3PfDJt7sxGIK9BsiEPL40s73X3S6+H1oSvpyODL7GUx6+q4WOvhqpnb4rA8K+6szHvjh6kL5ufKS+14T6vm8E/b4i+wG/WJUcv+11I7+2mS+/2JIcv2vUN7+t+jS/Fl03vz9pZr/lBFq/Md9Hv2DCbL9Wn4C/dJJ4vw==\",\"dtype\":\"float32\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"2868\"},\"selection_policy\":{\"id\":\"2867\"}},\"id\":\"2853\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00173\",\"sizing_mode\":\"stretch_width\"},\"id\":\"2885\",\"type\":\"Spacer\"},{\"attributes\":{\"children\":[{\"id\":\"2887\"},{\"id\":\"2925\"}],\"margin\":[0,0,0,0],\"name\":\"Column00204\"},\"id\":\"2886\",\"type\":\"Column\"},{\"attributes\":{\"source\":{\"id\":\"2853\"}},\"id\":\"2859\",\"type\":\"CDSView\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"2871\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"2870\",\"type\":\"Legend\"},{\"attributes\":{\"code\":\"def polynomial(\\n        coefficients: jnp.ndarray,\\n        n_observations: int = 100) -> [jnp.ndarray, jnp.ndarray]:\\n    x = jnp.linspace(0, 1, n_observations)\\n    y = jnp.polyval(coefficients, x)\\n    return x, \",\"filename\":\"\",\"height\":150,\"js_property_callbacks\":{\"change:disabled\":[{\"id\":\"2941\"}],\"change:readonly\":[{\"id\":\"2940\"}]},\"language\":\"python\",\"margin\":[5,10,5,10],\"sizing_mode\":\"stretch_both\",\"theme\":\"chaos\"},\"id\":\"2925\",\"type\":\"panel.models.ace.AcePlot\"},{\"attributes\":{\"children\":[{\"id\":\"2889\"},{\"id\":\"2890\"},{\"id\":\"2891\"},{\"id\":\"2892\"},{\"id\":\"2893\"},{\"id\":\"2894\"},{\"id\":\"2895\"},{\"id\":\"2896\"},{\"id\":\"2897\"}],\"margin\":[5,5,5,5],\"min_width\":300,\"name\":\"Problem parameters\",\"width\":300},\"id\":\"2888\",\"type\":\"Column\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"2858\"}]},\"id\":\"2871\",\"type\":\"LegendItem\"},{\"attributes\":{\"children\":[{\"id\":\"2888\"},{\"id\":\"2898\"}],\"margin\":[0,0,0,0],\"name\":\"Row00201\"},\"id\":\"2887\",\"type\":\"Row\"},{\"attributes\":{\"margin\":[5,10,5,10],\"name\":\"\",\"text\":\"<b>Problem parameters</b>\"},\"id\":\"2889\",\"type\":\"Div\"},{\"attributes\":{\"labels\":[\"Ground truth\"],\"margin\":[5,10,5,10]},\"id\":\"2892\",\"type\":\"CheckboxGroup\"},{\"attributes\":{\"labels\":[\"Error\"],\"margin\":[5,10,5,10]},\"id\":\"2893\",\"type\":\"CheckboxGroup\"},{\"attributes\":{\"end\":1000,\"margin\":[5,10,5,10],\"start\":10,\"title\":\"Number observations\",\"value\":100},\"id\":\"2890\",\"type\":\"Slider\"},{\"attributes\":{},\"id\":\"2832\",\"type\":\"BasicTicker\"},{\"attributes\":{\"end\":5.0,\"margin\":[5,10,5,10],\"start\":0.0,\"step\":0.1,\"title\":\"Noise level\",\"value\":0.1},\"id\":\"2891\",\"type\":\"Slider\"},{\"attributes\":{\"end\":5.0,\"margin\":[5,10,5,10],\"start\":-5.0,\"step\":0.1,\"title\":\"X 0\",\"value\":1.0},\"id\":\"2896\",\"type\":\"Slider\"},{\"attributes\":{\"end\":5.0,\"margin\":[5,10,5,10],\"start\":-5.0,\"step\":0.1,\"title\":\"X 2\",\"value\":-2.0},\"id\":\"2894\",\"type\":\"Slider\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"2874\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"2862\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"2911\"},\"renderers\":[{\"id\":\"2858\"}]},\"id\":\"2876\",\"type\":\"BoxSelectTool\"},{\"attributes\":{\"axis_label\":\"$$\\\\color{white} x$$\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"2865\"},\"group\":null,\"major_label_policy\":{\"id\":\"2866\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"2832\"}},\"id\":\"2831\",\"type\":\"LinearAxis\"},{\"attributes\":{\"end\":30,\"margin\":[5,10,5,10],\"start\":1,\"title\":\"Degree\",\"value\":1},\"id\":\"2897\",\"type\":\"Slider\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"2858\"}],\"tooltips\":null},\"id\":\"2872\",\"type\":\"HoverTool\"},{\"attributes\":{\"children\":[{\"id\":\"2820\"}],\"margin\":[0,0,0,0],\"name\":\"Row00198\"},\"id\":\"2898\",\"type\":\"Row\"},{\"attributes\":{\"line_alpha\":0.1,\"line_color\":\"red\",\"line_width\":4,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2880\",\"type\":\"Line\"},{\"attributes\":{\"line_alpha\":0.2,\"line_color\":\"red\",\"line_width\":4,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"2881\",\"type\":\"Line\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"2878\"},\"glyph\":{\"id\":\"2879\"},\"group\":null,\"hover_glyph\":null,\"muted_glyph\":{\"id\":\"2881\"},\"nonselection_glyph\":{\"id\":\"2880\"},\"view\":{\"id\":\"2883\"}},\"id\":\"2882\",\"type\":\"GlyphRenderer\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"2875\",\"type\":\"Circle\"}],\"root_ids\":[\"2884\",\"2968\"]},\"title\":\"Bokeh Application\",\"version\":\"2.4.3\"}};\n",
       "    var render_items = [{\"docid\":\"029f4470-e447-48bc-b39f-9c1cfce89c7e\",\"root_ids\":[\"2884\"],\"roots\":{\"2884\":\"b4045240-c9e1-44b9-86c6-e4ca34854d40\"}}];\n",
       "    root.Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "    for (const render_item of render_items) {\n",
       "      for (const root_id of render_item.root_ids) {\n",
       "\tconst id_el = document.getElementById(root_id)\n",
       "\tif (id_el.children.length && (id_el.children[0].className === 'bk-root')) {\n",
       "\t  const root_el = id_el.children[0]\n",
       "\t  root_el.id = root_el.id + '-rendered'\n",
       "\t}\n",
       "      }\n",
       "    }\n",
       "  }\n",
       "  if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "        clearInterval(timer);\n",
       "        embed_document(root);\n",
       "      } else if (document.readyState == \"complete\") {\n",
       "        attempts++;\n",
       "        if (attempts > 200) {\n",
       "          clearInterval(timer);\n",
       "          console.log(\"Bokeh: ERROR: Unable to run BokehJS code because BokehJS library is missing\");\n",
       "        }\n",
       "      }\n",
       "    }, 25, root)\n",
       "  }\n",
       "})(window);</script>"
      ],
      "text/plain": [
       "Row\n",
       "    [0] HSpacer()\n",
       "    [1] Column\n",
       "        [0] Row\n",
       "            [0] Param(RegressionPolynomialSolver, name='Problem parameters', widgets={})\n",
       "            [1] ParamMethod(method, _pane=Bokeh)\n",
       "        [1] Ace(height=150, language='python', sizing_mode='stretch_both', theme='chaos', value='def polynomial(\\n        ...)\n",
       "    [2] HSpacer()"
      ]
     },
     "execution_count": 15,
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {
       "id": "2884"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def solve(x, y, degree):\n",
    "    X = x[:, np.newaxis]\n",
    "    # Create our polynomial model for regression\n",
    "    model = make_pipeline(PolynomialFeatures(degree), LinearRegression())\n",
    "    # Fit the parameters of this model\n",
    "    model.fit(X, y)\n",
    "    # Predict the values\n",
    "    x_predict = np.linspace(np.min(x), np.max(x), 200)[:, np.newaxis]\n",
    "    y_model = model.predict(x_predict)\n",
    "    return x_predict[:, 0], y_model #np.array(jnp.zeros(y_model.shape))\n",
    "\n",
    "explorer.solve = solve\n",
    "explorer.render()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"linear\"></a>\n",
    "# Simple linear regression \n",
    "\n",
    "As discussed previously, regression allows to model the relationships that exist between inputs $\\mathbf{x}\\in\\mathbb{R}^{n}$ and a continuous output $y\\in\\mathbb{R}$. In the case of **linear** regression, we assume that the data that we observe comes from a linear relationship in the input, so that\n",
    "$$y = w_0 + w_1x_1 + w_2x_2 + ... + w_nx_n + \\epsilon$$\n",
    "with $\\epsilon$ exhibiting the *observation noise* (also called *residual error*) that is always present in measurements.\n",
    "\n",
    "To understand how we could learn a model approximating this relationship, we start with the case of *simple* linear regression, where $\\mathbf{x}\\in\\mathbb{R}$. This implies that our model will follow\n",
    "$$\\bar{y} = w_0 + w_1 x$$\n",
    "\n",
    "We will measure the errors made by our model by using the Mean Squared Error (MSE) loss, defined as \n",
    "$$\\mathcal{L}_{MSE}\\left( \\bar{\\mathbf{y}},\\theta \\right) = \\sum_{i=1}^{n} \\left| y_{i} - \\bar{y}_{i} \\right|^{2} = \\sum_{i=1}^{n} \\left| y_{i} - (w_{0} + w_{1} x_{i}) \\right|^{2}$$\n",
    "\n",
    "Then our goal is to find the most adequate set of parameters $\\theta = \\{w_{0}, w_{1}\\}$, which are those that minimize the MSE loss defined previously. Therefore, we aim to obtain\n",
    "$$\\theta^{*}= \\underset{\\theta}{\\text{argmin }} \\mathcal{L}\\left( \\bar{\\mathbf{y}}, \\mathbf{y} \\mid f_{\\theta}, \\theta \\right)$$\n",
    "\n",
    "To do so, we will implement the **gradient descent** algorithm discussed in the course.\n",
    "\n",
    "## Manual implementation - `NumPy`\n",
    "\n",
    "We start by performing a *full manual implementation*, in the sense that we need to manually derive the gradient in order to apply the gradient descent updates. To do so, we will rely on [NumPy](https://numpy.org/), which is a fundamental library for numerical computing offering support for N-dimensional arrays and scientific computing tasks, such as linear algebra, statistical analysis, and matrix manipulation. We strongly encourage you to learn NumPy through the set of [tutorials](https://numpy.org/learn/). For the sake of this introductory tutorial, we will provide the explanation for all of the functions that we will use in this first exercise. After that exercise, we will assume for the rest of the course that knowledge of Numpy should be found online. \n",
    "\n",
    "We start by import libraries (`NumPy`) and set a random seed to ensure that the random number generator produces always a reproducible series of random numbers.\n",
    "\n",
    "**Used functions**\n",
    "- `np.random.seed`: sets the seed for the NumPy random number generator. [Documentation](https://numpy.org/doc/stable/reference/random/generated/numpy.random.seed.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "np.random.seed(42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\u001b[0;31mDocstring:\u001b[0m\n",
       "seed(self, seed=None)\n",
       "\n",
       "Reseed a legacy MT19937 BitGenerator\n",
       "\n",
       "Notes\n",
       "-----\n",
       "This is a convenience, legacy function.\n",
       "\n",
       "The best practice is to **not** reseed a BitGenerator, rather to\n",
       "recreate a new one. This method is here for legacy reasons.\n",
       "This example demonstrates best practice.\n",
       "\n",
       ">>> from numpy.random import MT19937\n",
       ">>> from numpy.random import RandomState, SeedSequence\n",
       ">>> rs = RandomState(MT19937(SeedSequence(123456789)))\n",
       "# Later, you want to restart the stream\n",
       ">>> rs = RandomState(MT19937(SeedSequence(987654321)))\n",
       "\u001b[0;31mType:\u001b[0m      builtin_function_or_method"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "?np.random.seed"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Generate a synthetic dataset.\n",
    "\n",
    "For the sake of this exercise, we will generate the data ourselves, so that we know the true values that we are looking for in advance. Hence, we define a linear relationship following\n",
    "$$y = w^{t}_0 + w^{t}_1 x + \\epsilon$$\n",
    "\n",
    "**Used functions**\n",
    "- `np.random.rand`: generates an array of random numbers uniformly distributed over [0, 1). [Documentation](https://numpy.org/doc/stable/reference/random/generated/numpy.random.rand.html)\n",
    "- `np.random.randn`: generates an array of random numbers from the standard normal distribution (mean 0, variance 1). [Documentation](https://numpy.org/doc/stable/reference/random/generated/numpy.random.randn.html)\n",
    "- `np.ones`: generates an array of ones with a specified shape. [Documentation](https://numpy.org/doc/stable/reference/generated/numpy.ones.html)\n",
    "- `np.c_`: concatenates arrays along the second axis. [Documentation](https://numpy.org/doc/stable/reference/generated/numpy.c_.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "true_w0 = 2\n",
    "true_w1 = 3\n",
    "x_min = 0\n",
    "x_max = 1\n",
    "n_obs = 100\n",
    "x = np.linspace(x_min, x_max, n_obs)\n",
    "y = true_w1 * x + true_w0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Computing the MSE loss\n",
    "\n",
    "We will measure the errors made by our model by using the Mean Squared Error (MSE) loss, defined as \n",
    "$$\\mathcal{L}_{MSE}\\left( \\bar{\\mathbf{y}},\\theta \\right) = \\sum_{i=1}^{n} \\left| y_{i} - \\bar{y}_{i} \\right|^{2} = \\sum_{i=1}^{n} \\left| y_{i} - (w_{0} + w_{1} x_{i}) \\right|^{2}$$\n",
    "\n",
    "This will allow us to evaluate the performances of our model, but is also the basis for the following gradient descent algorithm.\n",
    "\n",
    "**Used functions**\n",
    "- `np.sum`: computes the sum of the provided array (optionally across a provided `axis`). [Documentation](https://numpy.org/doc/stable/reference/generated/numpy.mean.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def mse_loss(y, y_bar):\n",
    "    return np.sum((y - y_bar) ** 2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Implement the gradient descent algorithm.\n",
    "\n",
    "As seen in the course Initialize the weight $w_1$ and bias $w_0$ to small random values\n",
    "1. Evaluate the predictions made by our model \n",
    "$$\\hat{y} = w_{1}x + w_0$$\n",
    "2. Compute the mean squared error (MSE) loss between the predicted values and the ground truth labels: \n",
    "$$\\mathcal{L}_{MSE} = \\frac{1}{n} \\sum_{i=1}^n (\\bar{y_i} - y_i)^2$$\n",
    "3. Compute the gradients of the loss with respect to the weight and bias (see derivation in the course)\n",
    "$$\\frac{\\partial \\mathcal{L}}{\\partial w_1} = \\sum_{i} 2 * (\\bar{y}_{i} - y_{i})*x_{i}$$\n",
    "$$\\frac{\\partial \\mathcal{L}}{\\partial w_0} = \\sum_{i} 2 * (\\bar{y}_{i} - y_{i})$$\n",
    "4. Update the weights and bias using the gradients and a learning rate $\\eta$: \n",
    "$$w_1 \\leftarrow w_1 - \\eta \\frac{\\partial \\mathcal{L}}{\\partial w_1}$$ \n",
    "$$w_0 \\leftarrow w_0 - \\eta \\frac{\\partial \\mathcal{L}}{\\partial w_0}$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.999998555356199\n",
      "2.0000007742340746\n"
     ]
    }
   ],
   "source": [
    "# Parameters\n",
    "n_iter = 1000\n",
    "lr = 0.001\n",
    "def gradient_descent(x, y, n_iter, lr):\n",
    "    # Initialize the parameters\n",
    "    w_1 = np.random.randn()\n",
    "    w_0 = np.random.randn()\n",
    "    # Perform gradient descent\n",
    "    for i in range(n_iter):\n",
    "        # 1. Calculate the predictions\n",
    "        y_bar = w_1 * x + w_0\n",
    "        # 2. Compute the loss\n",
    "        loss = mse_loss(y, y_bar)\n",
    "        # 3. Calculate the gradients\n",
    "        dw_1 = 2 * np.sum((y_bar - y) * x)\n",
    "        dw_0 = 2 * np.sum((y_bar - y))\n",
    "        # 4. Update the parameters\n",
    "        w_1 -= lr * dw_1\n",
    "        w_0 -= lr * dw_0\n",
    "    return w_1, w_0\n",
    "w_1, w_0 = gradient_descent(x, y, n_iter, lr)\n",
    "print(w_1)\n",
    "print(w_0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualize the results.\n",
    "\n",
    "We can see how well our model fits to the observed data by plotting it against the observed samples."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.holoviews_exec.v0+json": "",
      "text/html": [
       "<div id='3112'>\n",
       "  <div class=\"bk-root\" id=\"2f548787-27ee-4c19-8a6e-3c4207e81198\" data-root-id=\"3112\"></div>\n",
       "</div>\n",
       "<script type=\"application/javascript\">(function(root) {\n",
       "  function embed_document(root) {\n",
       "    var docs_json = {\"0ad336bb-fc26-47fc-b6f2-77846a62cfd3\":{\"defs\":[{\"extends\":null,\"module\":null,\"name\":\"ReactiveHTML1\",\"overrides\":[],\"properties\":[]},{\"extends\":null,\"module\":null,\"name\":\"FlexBox1\",\"overrides\":[],\"properties\":[{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_content\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_items\"},{\"default\":\"row\",\"kind\":null,\"name\":\"flex_direction\"},{\"default\":\"wrap\",\"kind\":null,\"name\":\"flex_wrap\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"justify_content\"}]},{\"extends\":null,\"module\":null,\"name\":\"GridStack1\",\"overrides\":[],\"properties\":[{\"default\":\"warn\",\"kind\":null,\"name\":\"mode\"},{\"default\":null,\"kind\":null,\"name\":\"ncols\"},{\"default\":null,\"kind\":null,\"name\":\"nrows\"},{\"default\":true,\"kind\":null,\"name\":\"allow_resize\"},{\"default\":true,\"kind\":null,\"name\":\"allow_drag\"},{\"default\":[],\"kind\":null,\"name\":\"state\"}]},{\"extends\":null,\"module\":null,\"name\":\"click1\",\"overrides\":[],\"properties\":[{\"default\":\"\",\"kind\":null,\"name\":\"terminal_output\"},{\"default\":\"\",\"kind\":null,\"name\":\"debug_name\"},{\"default\":0,\"kind\":null,\"name\":\"clears\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationAreaBase1\",\"overrides\":[],\"properties\":[{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationArea1\",\"overrides\":[],\"properties\":[{\"default\":[],\"kind\":null,\"name\":\"notifications\"},{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"},{\"default\":[{\"background\":\"#ffc107\",\"icon\":{\"className\":\"fas fa-exclamation-triangle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"warning\"},{\"background\":\"#007bff\",\"icon\":{\"className\":\"fas fa-info-circle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"info\"}],\"kind\":null,\"name\":\"types\"}]},{\"extends\":null,\"module\":null,\"name\":\"Notification\",\"overrides\":[],\"properties\":[{\"default\":null,\"kind\":null,\"name\":\"background\"},{\"default\":3000,\"kind\":null,\"name\":\"duration\"},{\"default\":null,\"kind\":null,\"name\":\"icon\"},{\"default\":\"\",\"kind\":null,\"name\":\"message\"},{\"default\":null,\"kind\":null,\"name\":\"notification_type\"},{\"default\":false,\"kind\":null,\"name\":\"_destroyed\"}]},{\"extends\":null,\"module\":null,\"name\":\"TemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]},{\"extends\":null,\"module\":null,\"name\":\"MaterialTemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]}],\"roots\":{\"references\":[{\"attributes\":{\"axis_label\":\"x\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"3079\"},\"group\":null,\"major_label_policy\":{\"id\":\"3080\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"3046\"}},\"id\":\"3045\",\"type\":\"LinearAxis\"},{\"attributes\":{\"tools\":[{\"id\":\"3053\"},{\"id\":\"3054\"},{\"id\":\"3055\"},{\"id\":\"3056\"},{\"id\":\"3057\"},{\"id\":\"3058\"},{\"id\":\"3086\"},{\"id\":\"3090\"}]},\"id\":\"3060\",\"type\":\"Toolbar\"},{\"attributes\":{},\"id\":\"3082\",\"type\":\"Selection\"},{\"attributes\":{},\"id\":\"3041\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"3050\",\"type\":\"BasicTicker\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3071\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"3037\",\"type\":\"DataRange1d\"},{\"attributes\":{\"source\":{\"id\":\"3067\"}},\"id\":\"3073\",\"type\":\"CDSView\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"3085\"},{\"id\":\"3111\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"3084\",\"type\":\"Legend\"},{\"attributes\":{\"overlay\":{\"id\":\"3059\"}},\"id\":\"3055\",\"type\":\"BoxZoomTool\"},{\"attributes\":{},\"id\":\"3056\",\"type\":\"SaveTool\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"3045\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"3048\"},{\"id\":\"3052\"},{\"id\":\"3084\"}],\"height\":450,\"left\":[{\"id\":\"3049\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"3072\"},{\"id\":\"3096\"}],\"title\":{\"id\":\"3035\"},\"toolbar\":{\"id\":\"3060\"},\"toolbar_location\":null,\"x_range\":{\"id\":\"3037\"},\"x_scale\":{\"id\":\"3041\"},\"y_range\":{\"id\":\"3039\"},\"y_scale\":{\"id\":\"3043\"}},\"id\":\"3034\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"3088\",\"type\":\"Circle\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"3110\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"children\":[{\"id\":\"3113\"},{\"id\":\"3034\"},{\"id\":\"3140\"}],\"margin\":[0,0,0,0],\"name\":\"Row00213\"},\"id\":\"3112\",\"type\":\"Row\"},{\"attributes\":{\"axis\":{\"id\":\"3045\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"3048\",\"type\":\"Grid\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"3110\"},\"renderers\":[{\"id\":\"3072\"}]},\"id\":\"3090\",\"type\":\"BoxSelectTool\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00210\",\"sizing_mode\":\"stretch_width\"},\"id\":\"3140\",\"type\":\"Spacer\"},{\"attributes\":{},\"id\":\"3054\",\"type\":\"WheelZoomTool\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00209\",\"sizing_mode\":\"stretch_width\"},\"id\":\"3113\",\"type\":\"Spacer\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"3072\"}],\"tooltips\":null},\"id\":\"3086\",\"type\":\"HoverTool\"},{\"attributes\":{\"line_alpha\":0.1,\"line_color\":\"red\",\"line_width\":6,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3094\",\"type\":\"Line\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"3072\"}]},\"id\":\"3085\",\"type\":\"LegendItem\"},{\"attributes\":{\"axis\":{\"id\":\"3049\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"3052\",\"type\":\"Grid\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"3082\"},\"selection_policy\":{\"id\":\"3081\"}},\"id\":\"3067\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"line_alpha\":0.2,\"line_color\":\"red\",\"line_width\":6,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3095\",\"type\":\"Line\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"3067\"},\"glyph\":{\"id\":\"3068\"},\"group\":null,\"hover_glyph\":{\"id\":\"3070\"},\"muted_glyph\":{\"id\":\"3071\"},\"nonselection_glyph\":{\"id\":\"3089\"},\"selection_glyph\":{\"id\":\"3088\"},\"view\":{\"id\":\"3073\"}},\"id\":\"3072\",\"type\":\"GlyphRenderer\"},{\"attributes\":{},\"id\":\"3043\",\"type\":\"LinearScale\"},{\"attributes\":{\"source\":{\"id\":\"3092\"}},\"id\":\"3097\",\"type\":\"CDSView\"},{\"attributes\":{\"axis_label\":\"y\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"3076\"},\"group\":null,\"major_label_policy\":{\"id\":\"3077\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"3050\"}},\"id\":\"3049\",\"type\":\"LinearAxis\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"3089\",\"type\":\"Circle\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"3092\"},\"glyph\":{\"id\":\"3093\"},\"group\":null,\"hover_glyph\":null,\"muted_glyph\":{\"id\":\"3095\"},\"nonselection_glyph\":{\"id\":\"3094\"},\"view\":{\"id\":\"3097\"}},\"id\":\"3096\",\"type\":\"GlyphRenderer\"},{\"attributes\":{},\"id\":\"3076\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"label\":{\"value\":\"Learned model\"},\"renderers\":[{\"id\":\"3096\"}]},\"id\":\"3111\",\"type\":\"LegendItem\"},{\"attributes\":{\"line_alpha\":0.6,\"line_color\":\"red\",\"line_width\":6,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3093\",\"type\":\"Line\"},{\"attributes\":{},\"id\":\"3077\",\"type\":\"AllLabels\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3068\",\"type\":\"Circle\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"8XrqZwAAAEBqD9bpDz4AQOOjwWsffABAXDit7S66AEDUzJhvPvgAQE1hhPFNNgFAxvVvc110AUA/ilv1bLIBQLgeR3d88AFAMbMy+YsuAkCqRx57m2wCQCPcCf2qqgJAnHD1frroAkAUBeEAyiYDQI2ZzILZZANABi64BOmiA0B/wqOG+OADQPhWjwgIHwRAcet6ihddBEDqf2YMJ5sEQGIUUo422QRA26g9EEYXBUBUPSmSVVUFQM3RFBRlkwVARmYAlnTRBUC/+usXhA8GQDiP15mTTQZAsSPDG6OLBkAquK6dsskGQKJMmh/CBwdAG+GFodFFB0CUdXEj4YMHQA0KXaXwwQdAhp5IJwAACED/MjSpDz4IQHjHHysffAhA8FsLrS66CEBq8PYuPvgIQOKE4rBNNglAWxnOMl10CUDUrbm0bLIJQE1CpTZ88AlAxtaQuIsuCkA+a3w6m2wKQLj/Z7yqqgpAMJRTPrroCkCqKD/AySYLQCK9KkLZZAtAm1EWxOiiC0AU5gFG+OALQIx67ccHHwxABg/ZSRddDEB+o8TLJpsMQPc3sE022QxAcMybz0UXDUDpYIdRVVUNQGL1ctNkkw1A24leVXTRDUBUHkrXgw8OQMyyNVmTTQ5ARUch26KLDkC+2wxdsskOQDdw+N7BBw9AsATkYNFFD0Apmc/i4IMPQKItu2TwwQ9AG8Km5v//D0BKK0m0Bx8QQIb1PnUPPhBAwr80NhddEEAAiir3HnwQQDxUILgmmxBAeB4WeS66EEC06As6NtkQQPGyAfs9+BBALn33u0UXEUBqR+18TTYRQKYR4z1VVRFA4tvY/lx0EUAfps6/ZJMRQFxwxIBsshFAmDq6QXTREUDUBLACfPARQBHPpcODDxJATZmbhIsuEkCKY5FFk00SQMYthwabbBJAAvh8x6KLEkA/wnKIqqoSQHyMaEmyyRJAuFZeCrroEkD0IFTLwQcTQDHrSYzJJhNAbbU/TdFFE0CqfzUO2WQTQOZJK8/ggxNAIhQhkOiiE0Bf3hZR8METQJuoDBL44BNA2HIC0///E0A=\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"3108\"},\"selection_policy\":{\"id\":\"3107\"}},\"id\":\"3092\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Simple linear regression\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"3035\",\"type\":\"Title\"},{\"attributes\":{},\"id\":\"3057\",\"type\":\"ResetTool\"},{\"attributes\":{},\"id\":\"3053\",\"type\":\"PanTool\"},{\"attributes\":{},\"id\":\"3079\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{},\"id\":\"3046\",\"type\":\"BasicTicker\"},{\"attributes\":{},\"id\":\"3080\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"3058\",\"type\":\"HelpTool\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"3059\",\"type\":\"BoxAnnotation\"},{\"attributes\":{},\"id\":\"3107\",\"type\":\"UnionRenderers\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3070\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"3108\",\"type\":\"Selection\"},{\"attributes\":{\"client_comm_id\":\"ccd79467b2414417be150a1f958868a6\",\"comm_id\":\"88d55a05a4de4f5ebd573ee8370f97ef\",\"plot_id\":\"3112\"},\"id\":\"3180\",\"type\":\"panel.models.comm_manager.CommManager\"},{\"attributes\":{},\"id\":\"3039\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"3081\",\"type\":\"UnionRenderers\"}],\"root_ids\":[\"3112\",\"3180\"]},\"title\":\"Bokeh Application\",\"version\":\"2.4.3\"}};\n",
       "    var render_items = [{\"docid\":\"0ad336bb-fc26-47fc-b6f2-77846a62cfd3\",\"root_ids\":[\"3112\"],\"roots\":{\"3112\":\"2f548787-27ee-4c19-8a6e-3c4207e81198\"}}];\n",
       "    root.Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "    for (const render_item of render_items) {\n",
       "      for (const root_id of render_item.root_ids) {\n",
       "\tconst id_el = document.getElementById(root_id)\n",
       "\tif (id_el.children.length && (id_el.children[0].className === 'bk-root')) {\n",
       "\t  const root_el = id_el.children[0]\n",
       "\t  root_el.id = root_el.id + '-rendered'\n",
       "\t}\n",
       "      }\n",
       "    }\n",
       "  }\n",
       "  if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "        clearInterval(timer);\n",
       "        embed_document(root);\n",
       "      } else if (document.readyState == \"complete\") {\n",
       "        attempts++;\n",
       "        if (attempts > 200) {\n",
       "          clearInterval(timer);\n",
       "          console.log(\"Bokeh: ERROR: Unable to run BokehJS code because BokehJS library is missing\");\n",
       "        }\n",
       "      }\n",
       "    }, 25, root)\n",
       "  }\n",
       "})(window);</script>"
      ],
      "text/plain": [
       "Row\n",
       "    [0] HSpacer()\n",
       "    [1] Bokeh(Figure)\n",
       "    [2] HSpacer()"
      ]
     },
     "execution_count": 21,
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {
       "id": "3112"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p = scatter(x, y, title=\"Simple linear regression\")\n",
    "p.line(x, w_1 * x + w_0, line_width=6, line_alpha=0.6, color=\"red\", legend_label=r\"Learned model\")\n",
    "plot = (center_plot(p))\n",
    "plot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "TypeError: iteration over a 0-d array### Observing our solution interactively\n",
    "\n",
    "In the following code, you can observe the behavior of the model by playing interactively with the properties of the original problem. \n",
    "\n",
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #192841; border-color: #779ecb\">\n",
    "\n",
    "> Note that each change in the properties of the original problem requires to run the gradient descent algorithm entirely each time.\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.holoviews_exec.v0+json": "",
      "text/html": [
       "<div id='3324'>\n",
       "  <div class=\"bk-root\" id=\"9fdc3798-179f-4d5c-9124-f10897548316\" data-root-id=\"3324\"></div>\n",
       "</div>\n",
       "<script type=\"application/javascript\">(function(root) {\n",
       "  function embed_document(root) {\n",
       "    var docs_json = {\"924fd606-d4c6-4356-b6cd-327f4028862e\":{\"defs\":[{\"extends\":null,\"module\":null,\"name\":\"ReactiveHTML1\",\"overrides\":[],\"properties\":[]},{\"extends\":null,\"module\":null,\"name\":\"FlexBox1\",\"overrides\":[],\"properties\":[{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_content\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_items\"},{\"default\":\"row\",\"kind\":null,\"name\":\"flex_direction\"},{\"default\":\"wrap\",\"kind\":null,\"name\":\"flex_wrap\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"justify_content\"}]},{\"extends\":null,\"module\":null,\"name\":\"GridStack1\",\"overrides\":[],\"properties\":[{\"default\":\"warn\",\"kind\":null,\"name\":\"mode\"},{\"default\":null,\"kind\":null,\"name\":\"ncols\"},{\"default\":null,\"kind\":null,\"name\":\"nrows\"},{\"default\":true,\"kind\":null,\"name\":\"allow_resize\"},{\"default\":true,\"kind\":null,\"name\":\"allow_drag\"},{\"default\":[],\"kind\":null,\"name\":\"state\"}]},{\"extends\":null,\"module\":null,\"name\":\"click1\",\"overrides\":[],\"properties\":[{\"default\":\"\",\"kind\":null,\"name\":\"terminal_output\"},{\"default\":\"\",\"kind\":null,\"name\":\"debug_name\"},{\"default\":0,\"kind\":null,\"name\":\"clears\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationAreaBase1\",\"overrides\":[],\"properties\":[{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationArea1\",\"overrides\":[],\"properties\":[{\"default\":[],\"kind\":null,\"name\":\"notifications\"},{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"},{\"default\":[{\"background\":\"#ffc107\",\"icon\":{\"className\":\"fas fa-exclamation-triangle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"warning\"},{\"background\":\"#007bff\",\"icon\":{\"className\":\"fas fa-info-circle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"info\"}],\"kind\":null,\"name\":\"types\"}]},{\"extends\":null,\"module\":null,\"name\":\"Notification\",\"overrides\":[],\"properties\":[{\"default\":null,\"kind\":null,\"name\":\"background\"},{\"default\":3000,\"kind\":null,\"name\":\"duration\"},{\"default\":null,\"kind\":null,\"name\":\"icon\"},{\"default\":\"\",\"kind\":null,\"name\":\"message\"},{\"default\":null,\"kind\":null,\"name\":\"notification_type\"},{\"default\":false,\"kind\":null,\"name\":\"_destroyed\"}]},{\"extends\":null,\"module\":null,\"name\":\"TemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]},{\"extends\":null,\"module\":null,\"name\":\"MaterialTemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]}],\"roots\":{\"references\":[{\"attributes\":{\"line_alpha\":0.1,\"line_color\":\"red\",\"line_width\":4,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3306\",\"type\":\"Line\"},{\"attributes\":{},\"id\":\"3293\",\"type\":\"UnionRenderers\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"3271\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"3279\"},\"glyph\":{\"id\":\"3280\"},\"group\":null,\"hover_glyph\":{\"id\":\"3282\"},\"muted_glyph\":{\"id\":\"3283\"},\"nonselection_glyph\":{\"id\":\"3301\"},\"selection_glyph\":{\"id\":\"3300\"},\"view\":{\"id\":\"3285\"}},\"id\":\"3284\",\"type\":\"GlyphRenderer\"},{\"attributes\":{},\"id\":\"3294\",\"type\":\"Selection\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"3304\"},\"glyph\":{\"id\":\"3305\"},\"group\":null,\"hover_glyph\":null,\"muted_glyph\":{\"id\":\"3307\"},\"nonselection_glyph\":{\"id\":\"3306\"},\"view\":{\"id\":\"3309\"}},\"id\":\"3308\",\"type\":\"GlyphRenderer\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3283\",\"type\":\"Circle\"},{\"attributes\":{\"tools\":[{\"id\":\"3265\"},{\"id\":\"3266\"},{\"id\":\"3267\"},{\"id\":\"3268\"},{\"id\":\"3269\"},{\"id\":\"3270\"},{\"id\":\"3298\"},{\"id\":\"3302\"}]},\"id\":\"3272\",\"type\":\"Toolbar\"},{\"attributes\":{\"source\":{\"id\":\"3304\"}},\"id\":\"3309\",\"type\":\"CDSView\"},{\"attributes\":{\"line_alpha\":0.2,\"line_color\":\"red\",\"line_width\":4,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3307\",\"type\":\"Line\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3282\",\"type\":\"Circle\"},{\"attributes\":{\"source\":{\"id\":\"3279\"}},\"id\":\"3285\",\"type\":\"CDSView\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"3297\"},{\"id\":\"3323\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"3296\",\"type\":\"Legend\"},{\"attributes\":{\"args\":{\"bidirectional\":true,\"properties\":{\"readonly\":\"disabled\"},\"source\":{\"id\":\"3363\"},\"target\":{\"id\":\"3363\"}},\"code\":\"\\n    var value = source['readonly'];\\n    value = value;\\n    value = value;\\n    try {\\n      var property = target.properties['disabled'];\\n      if (property !== undefined) { property.validate(value); }\\n    } catch(err) {\\n      console.log('WARNING: Could not set disabled on target, raised error: ' + err);\\n      return;\\n    }\\n    try {\\n      target['disabled'] = value;\\n    } catch(err) {\\n      console.log(err)\\n    }\\n    \",\"tags\":[[140461253792048,[null,\"readonly\"],[null,\"disabled\"]]]},\"id\":\"3378\",\"type\":\"CustomJS\"},{\"attributes\":{\"children\":[{\"id\":\"3325\"},{\"id\":\"3326\"},{\"id\":\"3364\"}],\"margin\":[0,0,0,0],\"name\":\"Row00246\"},\"id\":\"3324\",\"type\":\"Row\"},{\"attributes\":{\"client_comm_id\":\"fde6d35f64fd4db8bfedfaeb8a7c7942\",\"comm_id\":\"079417b2d54b455eb301eb333eda8bad\",\"plot_id\":\"3324\"},\"id\":\"3406\",\"type\":\"panel.models.comm_manager.CommManager\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00217\",\"sizing_mode\":\"stretch_width\"},\"id\":\"3325\",\"type\":\"Spacer\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"3284\"}]},\"id\":\"3297\",\"type\":\"LegendItem\"},{\"attributes\":{\"label\":{\"value\":\"Learned model\"},\"renderers\":[{\"id\":\"3308\"}]},\"id\":\"3323\",\"type\":\"LegendItem\"},{\"attributes\":{\"children\":[{\"id\":\"3327\"},{\"id\":\"3363\"}],\"margin\":[0,0,0,0],\"name\":\"Column00244\"},\"id\":\"3326\",\"type\":\"Column\"},{\"attributes\":{\"args\":{\"bidirectional\":true,\"properties\":{\"readonly\":\"disabled\"},\"source\":{\"id\":\"3363\"},\"target\":{\"id\":\"3363\"}},\"code\":\"\\n    var value = source['disabled'];\\n    value = value;\\n    value = value;\\n    try {\\n      var property = target.properties['readonly'];\\n      if (property !== undefined) { property.validate(value); }\\n    } catch(err) {\\n      console.log('WARNING: Could not set readonly on target, raised error: ' + err);\\n      return;\\n    }\\n    try {\\n      target['readonly'] = value;\\n    } catch(err) {\\n      console.log(err)\\n    }\\n    \",\"tags\":[[140461253792048,[null,\"readonly\"],[null,\"disabled\"]]]},\"id\":\"3379\",\"type\":\"CustomJS\"},{\"attributes\":{\"children\":[{\"id\":\"3329\"},{\"id\":\"3330\"},{\"id\":\"3331\"},{\"id\":\"3332\"},{\"id\":\"3333\"},{\"id\":\"3334\"},{\"id\":\"3335\"}],\"margin\":[5,5,5,5],\"min_width\":300,\"name\":\"Problem parameters\",\"width\":300},\"id\":\"3328\",\"type\":\"Column\"},{\"attributes\":{\"children\":[{\"id\":\"3328\"},{\"id\":\"3336\"}],\"margin\":[0,0,0,0],\"name\":\"Row00241\"},\"id\":\"3327\",\"type\":\"Row\"},{\"attributes\":{\"margin\":[5,10,5,10],\"name\":\"\",\"text\":\"<b>Problem parameters</b>\"},\"id\":\"3329\",\"type\":\"Div\"},{\"attributes\":{\"labels\":[\"Ground truth\"],\"margin\":[5,10,5,10]},\"id\":\"3332\",\"type\":\"CheckboxGroup\"},{\"attributes\":{},\"id\":\"3265\",\"type\":\"PanTool\"},{\"attributes\":{\"labels\":[\"Error\"],\"margin\":[5,10,5,10]},\"id\":\"3333\",\"type\":\"CheckboxGroup\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3280\",\"type\":\"Circle\"},{\"attributes\":{\"end\":1000,\"margin\":[5,10,5,10],\"start\":10,\"title\":\"Number observations\",\"value\":100},\"id\":\"3330\",\"type\":\"Slider\"},{\"attributes\":{\"end\":5.0,\"margin\":[5,10,5,10],\"start\":0.0,\"step\":0.1,\"title\":\"Noise level\",\"value\":0.1},\"id\":\"3331\",\"type\":\"Slider\"},{\"attributes\":{},\"id\":\"3291\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"end\":5.0,\"margin\":[5,10,5,10],\"start\":-5.0,\"step\":0.1,\"title\":\"X 1\",\"value\":1.0},\"id\":\"3334\",\"type\":\"Slider\"},{\"attributes\":{},\"id\":\"3292\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"3319\",\"type\":\"UnionRenderers\"},{\"attributes\":{\"code\":\"def linear(\\n        coefficients: jnp.ndarray,\\n        n_observations: int = 100) -> [jnp.ndarray, jnp.ndarray]:\\n    x = jnp.linspace(0, 1, n_observations)\\n    y = coefficients[0] * x + coefficients[1]\\n    return x, \",\"filename\":\"\",\"height\":150,\"js_property_callbacks\":{\"change:disabled\":[{\"id\":\"3379\"}],\"change:readonly\":[{\"id\":\"3378\"}]},\"language\":\"python\",\"margin\":[5,10,5,10],\"sizing_mode\":\"stretch_both\",\"theme\":\"chaos\"},\"id\":\"3363\",\"type\":\"panel.models.ace.AcePlot\"},{\"attributes\":{},\"id\":\"3258\",\"type\":\"BasicTicker\"},{\"attributes\":{\"end\":5.0,\"margin\":[5,10,5,10],\"start\":-5.0,\"step\":0.1,\"title\":\"X 0\",\"value\":0.0},\"id\":\"3335\",\"type\":\"Slider\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"3257\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"3260\"},{\"id\":\"3264\"},{\"id\":\"3296\"}],\"height\":450,\"left\":[{\"id\":\"3261\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"3284\"},{\"id\":\"3308\"}],\"title\":{\"id\":\"3247\"},\"toolbar\":{\"id\":\"3272\"},\"toolbar_location\":null,\"x_range\":{\"id\":\"3249\"},\"x_scale\":{\"id\":\"3253\"},\"y_range\":{\"id\":\"3251\"},\"y_scale\":{\"id\":\"3255\"}},\"id\":\"3246\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{\"children\":[{\"id\":\"3246\"}],\"margin\":[0,0,0,0],\"name\":\"Row00238\"},\"id\":\"3336\",\"type\":\"Row\"},{\"attributes\":{},\"id\":\"3320\",\"type\":\"Selection\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Linear regression problem\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"3247\",\"type\":\"Title\"},{\"attributes\":{},\"id\":\"3255\",\"type\":\"LinearScale\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00245\",\"sizing_mode\":\"stretch_width\"},\"id\":\"3364\",\"type\":\"Spacer\"},{\"attributes\":{\"axis_label\":\"$$\\\\color{white} x$$\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"3291\"},\"group\":null,\"major_label_policy\":{\"id\":\"3292\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"3258\"}},\"id\":\"3257\",\"type\":\"LinearAxis\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"3300\",\"type\":\"Circle\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"3322\"},\"renderers\":[{\"id\":\"3284\"}]},\"id\":\"3302\",\"type\":\"BoxSelectTool\"},{\"attributes\":{},\"id\":\"3289\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"3249\",\"type\":\"DataRange1d\"},{\"attributes\":{\"line_alpha\":0.7,\"line_color\":\"red\",\"line_width\":4,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3305\",\"type\":\"Line\"},{\"attributes\":{\"axis_label\":\"$$\\\\color{white} 1.0x^1 + 0.0$$\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"3288\"},\"group\":null,\"major_label_policy\":{\"id\":\"3289\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"3262\"}},\"id\":\"3261\",\"type\":\"LinearAxis\"},{\"attributes\":{},\"id\":\"3269\",\"type\":\"ResetTool\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"3284\"}],\"tooltips\":null},\"id\":\"3298\",\"type\":\"HoverTool\"},{\"attributes\":{},\"id\":\"3251\",\"type\":\"DataRange1d\"},{\"attributes\":{},\"id\":\"3253\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"3288\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"3301\",\"type\":\"Circle\"},{\"attributes\":{\"axis\":{\"id\":\"3257\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"3260\",\"type\":\"Grid\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"3322\",\"type\":\"BoxAnnotation\"},{\"attributes\":{\"axis\":{\"id\":\"3261\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"3264\",\"type\":\"Grid\"},{\"attributes\":{},\"id\":\"3262\",\"type\":\"BasicTicker\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float32\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float32\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"3294\"},\"selection_policy\":{\"id\":\"3293\"}},\"id\":\"3279\",\"type\":\"ColumnDataSource\"},{\"attributes\":{},\"id\":\"3270\",\"type\":\"HelpTool\"},{\"attributes\":{},\"id\":\"3266\",\"type\":\"WheelZoomTool\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[200]},\"y\":{\"__ndarray__\":\"wOzbu+yS3LqURls7qDUJPKqZWzzX/pY82DDAPNhi6TxtSgk9buMdPW98Mj1vFUc9cK5bPXBHcD04cII9ubyMPTkJlz25VaE9OqKrPbrutT07O8A9u4fKPTvU1D28IN89PG3pPby58z08Bv49XykEPp5PCT7fdQ4+H5wTPl/CGD6f6B0+4A4jPh81KD5gWy0+oIEyPt+nNz4gzjw+YPRBPqEaRz7gQEw+IWdRPmGNVj6hs1s+4dlgPiIAZj5hJms+okxwPuJydT4imXo+Yr9/PtFygj7yBYU+EpmHPjEsij5Rv4w+cVKPPpLlkT6yeJQ+0guXPvGemT4SMpw+MsWePlJYoT5y66M+k36mPrMRqT7SpKs+8jeuPhPLsD4zXrM+U/G1PnOEuD6SF7s+s6q9PtM9wD7z0MI+E2TFPjT3xz5Uiso+cx3NPpOwzz60Q9I+1NbUPvRp1z4U/dk+NJDcPlQj3z50tuE+lEnkPrTc5j7Vb+k+9QLsPhSW7j40KfE+VbzzPnVP9j6V4vg+tXX7PtUI/j76TQA/ipcBPxrhAj+qKgQ/OnQFP8y9Bj9aBwg/7FAJP3qaCj8K5As/nC0NPyp3Dj+8wA8/SgoRP9xTEj9snRM/+uYUP4wwFj8cehc/rMMYPzwNGj/KVhs/XKAcP+zpHT98Mx8/DH0gP5zGIT8sECM/vFkkP0yjJT/c7CY/bDYoP/x/KT+MySo/HBMsP6xcLT88pi4/zO8vP1w5MT/sgjI/fMwzPwwWNT+cXzY/LKk3P7zyOD9MPDo/3IU7P2zPPD/8GD4/jGI/PxysQD+s9UE/Pj9DP8yIRD9c0kU/7BtHP3xlSD8Or0k/nPhKPy5CTD++i00/TNVOP94eUD9saFE//rFSP477Uz8cRVU/ro5WPzzYVz/OIVk/XmtaP+60Wz9+/lw/DkheP56RXz8u22A/viRiP05uYz/et2Q/bgFmP/5KZz+OlGg/Ht5pP64naz8+cWw/zrptP14Ebz/uTXA/fpdxPw7hcj+eKnQ/LnR1P769dj9OB3g/3lB5P26aej/+43s/ji19Px53fj8=\",\"dtype\":\"float32\",\"order\":\"little\",\"shape\":[200]}},\"selected\":{\"id\":\"3320\"},\"selection_policy\":{\"id\":\"3319\"}},\"id\":\"3304\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"overlay\":{\"id\":\"3271\"}},\"id\":\"3267\",\"type\":\"BoxZoomTool\"},{\"attributes\":{},\"id\":\"3268\",\"type\":\"SaveTool\"}],\"root_ids\":[\"3324\",\"3406\"]},\"title\":\"Bokeh Application\",\"version\":\"2.4.3\"}};\n",
       "    var render_items = [{\"docid\":\"924fd606-d4c6-4356-b6cd-327f4028862e\",\"root_ids\":[\"3324\"],\"roots\":{\"3324\":\"9fdc3798-179f-4d5c-9124-f10897548316\"}}];\n",
       "    root.Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "    for (const render_item of render_items) {\n",
       "      for (const root_id of render_item.root_ids) {\n",
       "\tconst id_el = document.getElementById(root_id)\n",
       "\tif (id_el.children.length && (id_el.children[0].className === 'bk-root')) {\n",
       "\t  const root_el = id_el.children[0]\n",
       "\t  root_el.id = root_el.id + '-rendered'\n",
       "\t}\n",
       "      }\n",
       "    }\n",
       "  }\n",
       "  if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "        clearInterval(timer);\n",
       "        embed_document(root);\n",
       "      } else if (document.readyState == \"complete\") {\n",
       "        attempts++;\n",
       "        if (attempts > 200) {\n",
       "          clearInterval(timer);\n",
       "          console.log(\"Bokeh: ERROR: Unable to run BokehJS code because BokehJS library is missing\");\n",
       "        }\n",
       "      }\n",
       "    }, 25, root)\n",
       "  }\n",
       "})(window);</script>"
      ],
      "text/plain": [
       "Row\n",
       "    [0] HSpacer()\n",
       "    [1] Column\n",
       "        [0] Row\n",
       "            [0] Param(RegressionLinearSolver, name='Problem parameters', widgets={})\n",
       "            [1] ParamMethod(method, _pane=Bokeh)\n",
       "        [1] Ace(height=150, language='python', sizing_mode='stretch_both', theme='chaos', value='def linear(\\n        coef...)\n",
       "    [2] HSpacer()"
      ]
     },
     "execution_count": 22,
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {
       "id": "3324"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from cml.tasks import RegressionLinearSolver\n",
    "explorer = RegressionLinearSolver()\n",
    "def solve(x, y):\n",
    "    w_1, w_0 = gradient_descent(x, y, n_iter, lr)\n",
    "    x_predict = np.linspace(np.min(x), np.max(x), 200)\n",
    "    y_model = w_1 * x_predict + w_0\n",
    "    return np.array(x_predict), np.array(y_model) #np.array(np.zeros(y_model.shape))\n",
    "\n",
    "explorer.solve = solve\n",
    "explorer.render()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #192841; border-color: #779ecb\">\n",
    "\n",
    "> ### Going further\n",
    "> **Exercise 1 \\[Learning rate\\]**: Experiment with different learning rates (e.g., 0.001, 0.01, 0.1, 1) and observe how they affect the convergence of the gradient descent algorithm. Plot the convergence history (cost function value vs. iteration) for each learning rate.\n",
    "\n",
    "> **Exercise 2 \\[Initialization\\]**: Experiment with different initializations of the coefficients (zeros, random values, etc.) and observe their impact on the convergence of the gradient descent algorithm.\n",
    "\n",
    "> **Exercise 3 \\[Variants\\]**: Implement and compare different variants of gradient descent, such as stochastic gradient descent (SGD) and mini-batch gradient descent. Analyze their convergence properties and computational efficiency.\n",
    "\n",
    "> **Exercise 4 \\[Regularization\\]**: Implement Lasso and Ridge regression with gradient descent by incorporating the regularization terms into the cost function and gradient calculations. Compare their performance with the standard linear regression implementation and analyze their impact on the learned coefficients.\n",
    "\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Discovering automatic differentiation - `JAX`\n",
    "\n",
    "The previous implementation required us to perform manual differentiation of our loss function to understand how to update the parameters. However, large developments have been made in the field of **automatic differentiation**. \n",
    "\n",
    "The recent library [JAX](https://github.com/google/jax) extends NumPy with this automatic differentiation feature (*autograd*), while providing a functional approach to numerical computing, allowing for easy gradient computation and just-in-time (JIT) compilation. Its ability to handle complex and custom gradients makes JAX particularly well-suited for advanced research projects. Similar to NumPy, we strongly encourage you to learn JAX through the set of [tutorials](https://jax.readthedocs.io/en/latest/), but we will also provide here the explanation for all of the functions that we will use in this first exercise. After that, we will assume that knowledge of JAX should be found online. \n",
    "\n",
    "Note that `JAX` has been thought as an extension of `NumPy`, therefore an extremely large portion of its API simply mirrors the `NumPy` functions by adding automatic differentiation features to it.\n",
    "\n",
    "**Used functions**\n",
    "- `random.PRNGKey`: function to generate a key for the pseudorandom number generator. [Documentation](https://jax.readthedocs.io/en/latest/_autosummary/jax.random.PRNGKey.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import jax.numpy as jnp\n",
    "from jax import grad, jit, random\n",
    "key = random.PRNGKey(42)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Dataset.\n",
    "\n",
    "We can simply keep our previous approach to generating the dataset but we will rely on `jnp.ndarray` instead of `np.ndarray`. We also provide the code for splitting the dataset between a `training` and `validation` dataset, as discussed in the course\n",
    "\n",
    "**Used functions**\n",
    "- `random.split`: function to split a PRNGKey into a list of subkeys [Documentation](https://jax.readthedocs.io/en/latest/_autosummary/jax.random.split.html)\n",
    "- `random.uniform`: function to generate an array of uniformly distributed random numbers [Documentation](https://jax.readthedocs.io/en/latest/_autosummary/jax.random.uniform.html)\n",
    "- `random.normal`: function to generate an array of normally distributed random numbers [Documentation](https://jax.readthedocs.io/en/latest/_autosummary/jax.random.normal.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "true_w0 = 2\n",
    "true_w1 = 3\n",
    "key_0, key_1, key = random.split(key, 3)\n",
    "x = random.uniform(key_0, (100,))\n",
    "y = true_w0 + true_w1 * x + (random.normal(key_1, (100,)) * 0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Splitting our dataset\n",
    "train_size = int(0.8 * len(x))\n",
    "x_train, x_valid = x[:train_size], x[train_size:]\n",
    "y_train, y_valid = y[:train_size], y[train_size:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Loss function\n",
    "\n",
    "Similarly Define the cost function and its gradient.\n",
    "\n",
    "**Used functions**\n",
    "- `jnp.mean`: function to compute the mean of an array [Documentation](https://jax.readthedocs.io/en/latest/_autosummary/jax.numpy.mean.html)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def loss_function(x, y, w_1, w_0):\n",
    "    y_pred = x * w_1 + w_0\n",
    "    residuals = y_pred - y\n",
    "    print(residuals)\n",
    "    return jnp.mean(residuals**2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "**Used functions**\n",
    "- `jit`: function to compile a function for faster execution [Documentation](https://jax.readthedocs.io/en/latest/jit.html)\n",
    "- `grad`: function to compute the gradient of a function [Documentation](https://jax.readthedocs.io/en/latest/jax.html#jax.grad)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "grad_loss_function = jit(grad(loss_function, argnums=[2, 3]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Implement the gradient descent algorithm.\n",
    "\n",
    "Note that nowhere in the code do we need to explicitly define the gradients of the different variables. Yet, the optimization will be performed adequately."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def gradient_descent(key, x, y, lr=0.05, n_iter=1000):\n",
    "    m = x.shape\n",
    "    k_0, k_1 = random.split(key, 2)\n",
    "    w_0 = random.normal(k_0, (1,))\n",
    "    w_1 = random.normal(k_1, (1,))\n",
    "    loss_history = []\n",
    "    for _ in range(n_iter):\n",
    "        gradients = grad_loss_function(x, y, w_1, w_0)\n",
    "        w_1 -= lr * gradients[0]\n",
    "        w_0 -= lr * gradients[1]\n",
    "        loss_history.append(loss_function(x, y, w_1, w_0))\n",
    "    return w_1, w_0, loss_history\n",
    "\n",
    "key_gd, key = random.split(key, 2)\n",
    "w_1, w_0, loss_history = gradient_descent(key_gd, x_train, y_train)\n",
    "print(\"Gradient Descent coefficients:\", w_1, w_0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluate our performances\n",
    "\n",
    "Thanks to our previous split of the dataset, we are now able to evaluate the performances of our model on *unseen data*, which is the overarching goal of building such machine learning models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE for Gradient Descent: 0.03851322\n"
     ]
    }
   ],
   "source": [
    "def mean_squared_error(y_true, y_pred):\n",
    "    return np.mean((y_true - y_pred)**2)\n",
    "\n",
    "y_valid_gd = w_1 * x_valid + w_0\n",
    "mse_gd = mean_squared_error(y_valid, y_valid_gd)\n",
    "print(\"MSE for Gradient Descent:\", mse_gd)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualize the results.\n",
    "\n",
    "As previously we can witness our results, and also evaluate our solution with our interactive solver."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.holoviews_exec.v0+json": "",
      "text/html": [
       "<div id='3550'>\n",
       "  <div class=\"bk-root\" id=\"5bb4e4b9-9773-4ba6-ae64-1832b032611c\" data-root-id=\"3550\"></div>\n",
       "</div>\n",
       "<script type=\"application/javascript\">(function(root) {\n",
       "  function embed_document(root) {\n",
       "    var docs_json = {\"3947ceaa-53d1-484c-b22b-cd9e2db43368\":{\"defs\":[{\"extends\":null,\"module\":null,\"name\":\"ReactiveHTML1\",\"overrides\":[],\"properties\":[]},{\"extends\":null,\"module\":null,\"name\":\"FlexBox1\",\"overrides\":[],\"properties\":[{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_content\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_items\"},{\"default\":\"row\",\"kind\":null,\"name\":\"flex_direction\"},{\"default\":\"wrap\",\"kind\":null,\"name\":\"flex_wrap\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"justify_content\"}]},{\"extends\":null,\"module\":null,\"name\":\"GridStack1\",\"overrides\":[],\"properties\":[{\"default\":\"warn\",\"kind\":null,\"name\":\"mode\"},{\"default\":null,\"kind\":null,\"name\":\"ncols\"},{\"default\":null,\"kind\":null,\"name\":\"nrows\"},{\"default\":true,\"kind\":null,\"name\":\"allow_resize\"},{\"default\":true,\"kind\":null,\"name\":\"allow_drag\"},{\"default\":[],\"kind\":null,\"name\":\"state\"}]},{\"extends\":null,\"module\":null,\"name\":\"click1\",\"overrides\":[],\"properties\":[{\"default\":\"\",\"kind\":null,\"name\":\"terminal_output\"},{\"default\":\"\",\"kind\":null,\"name\":\"debug_name\"},{\"default\":0,\"kind\":null,\"name\":\"clears\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationAreaBase1\",\"overrides\":[],\"properties\":[{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationArea1\",\"overrides\":[],\"properties\":[{\"default\":[],\"kind\":null,\"name\":\"notifications\"},{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"},{\"default\":[{\"background\":\"#ffc107\",\"icon\":{\"className\":\"fas fa-exclamation-triangle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"warning\"},{\"background\":\"#007bff\",\"icon\":{\"className\":\"fas fa-info-circle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"info\"}],\"kind\":null,\"name\":\"types\"}]},{\"extends\":null,\"module\":null,\"name\":\"Notification\",\"overrides\":[],\"properties\":[{\"default\":null,\"kind\":null,\"name\":\"background\"},{\"default\":3000,\"kind\":null,\"name\":\"duration\"},{\"default\":null,\"kind\":null,\"name\":\"icon\"},{\"default\":\"\",\"kind\":null,\"name\":\"message\"},{\"default\":null,\"kind\":null,\"name\":\"notification_type\"},{\"default\":false,\"kind\":null,\"name\":\"_destroyed\"}]},{\"extends\":null,\"module\":null,\"name\":\"TemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]},{\"extends\":null,\"module\":null,\"name\":\"MaterialTemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]}],\"roots\":{\"references\":[{\"attributes\":{\"line_alpha\":0.2,\"line_color\":\"red\",\"line_width\":6,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3533\",\"type\":\"Line\"},{\"attributes\":{},\"id\":\"3514\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{},\"id\":\"3515\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"3517\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"axis_label\":\"x\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"3517\"},\"group\":null,\"major_label_policy\":{\"id\":\"3518\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"3484\"}},\"id\":\"3483\",\"type\":\"LinearAxis\"},{\"attributes\":{},\"id\":\"3481\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"3518\",\"type\":\"AllLabels\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"3497\",\"type\":\"BoxAnnotation\"},{\"attributes\":{},\"id\":\"3494\",\"type\":\"SaveTool\"},{\"attributes\":{\"axis\":{\"id\":\"3483\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"3486\",\"type\":\"Grid\"},{\"attributes\":{\"axis_label\":\"y\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"3514\"},\"group\":null,\"major_label_policy\":{\"id\":\"3515\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"3488\"}},\"id\":\"3487\",\"type\":\"LinearAxis\"},{\"attributes\":{\"line_alpha\":0.6,\"line_color\":\"red\",\"line_width\":6,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3531\",\"type\":\"Line\"},{\"attributes\":{\"source\":{\"id\":\"3505\"}},\"id\":\"3511\",\"type\":\"CDSView\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float32\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float32\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"3546\"},\"selection_policy\":{\"id\":\"3545\"}},\"id\":\"3530\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.25},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.25},\"line_alpha\":{\"value\":0.25},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3506\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"3491\",\"type\":\"PanTool\"},{\"attributes\":{\"label\":{\"value\":\"Observations\"},\"renderers\":[{\"id\":\"3510\"}]},\"id\":\"3523\",\"type\":\"LegendItem\"},{\"attributes\":{\"source\":{\"id\":\"3530\"}},\"id\":\"3535\",\"type\":\"CDSView\"},{\"attributes\":{\"label\":{\"value\":\"Learned model\"},\"renderers\":[{\"id\":\"3534\"}]},\"id\":\"3549\",\"type\":\"LegendItem\"},{\"attributes\":{},\"id\":\"3519\",\"type\":\"UnionRenderers\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"3483\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"3486\"},{\"id\":\"3490\"},{\"id\":\"3522\"}],\"height\":450,\"left\":[{\"id\":\"3487\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"3510\"},{\"id\":\"3534\"}],\"title\":{\"id\":\"3473\"},\"toolbar\":{\"id\":\"3498\"},\"toolbar_location\":null,\"x_range\":{\"id\":\"3475\"},\"x_scale\":{\"id\":\"3479\"},\"y_range\":{\"id\":\"3477\"},\"y_scale\":{\"id\":\"3481\"}},\"id\":\"3472\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{\"children\":[{\"id\":\"3551\"},{\"id\":\"3472\"},{\"id\":\"3578\"}],\"margin\":[0,0,0,0],\"name\":\"Row00253\"},\"id\":\"3550\",\"type\":\"Row\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"value\":\"midnightblue\"},\"hatch_alpha\":{\"value\":0.2},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3509\",\"type\":\"Circle\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00250\",\"sizing_mode\":\"stretch_width\"},\"id\":\"3578\",\"type\":\"Spacer\"},{\"attributes\":{},\"id\":\"3520\",\"type\":\"Selection\"},{\"attributes\":{},\"id\":\"3475\",\"type\":\"DataRange1d\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00249\",\"sizing_mode\":\"stretch_width\"},\"id\":\"3551\",\"type\":\"Spacer\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"3548\",\"type\":\"BoxAnnotation\"},{\"attributes\":{},\"id\":\"3477\",\"type\":\"DataRange1d\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Linear regression\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"3473\",\"type\":\"Title\"},{\"attributes\":{},\"id\":\"3545\",\"type\":\"UnionRenderers\"},{\"attributes\":{\"line_alpha\":0.1,\"line_color\":\"red\",\"line_width\":6,\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3532\",\"type\":\"Line\"},{\"attributes\":{\"client_comm_id\":\"09d250d368e2442ea3f1981cce6525c0\",\"comm_id\":\"dd395ec88b9c4f4dba2cbdc7fd74a865\",\"plot_id\":\"3550\"},\"id\":\"3618\",\"type\":\"panel.models.comm_manager.CommManager\"},{\"attributes\":{\"overlay\":{\"id\":\"3497\"}},\"id\":\"3493\",\"type\":\"BoxZoomTool\"},{\"attributes\":{\"background_fill_alpha\":0.25,\"background_fill_color\":\"#505050\",\"border_line_alpha\":0.25,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"3523\"},{\"id\":\"3549\"}],\"label_standoff\":8,\"label_text_color\":\"#E0E0E0\",\"label_text_font\":\"Josefin Sans\",\"label_text_font_size\":\"1.15em\",\"spacing\":8},\"id\":\"3522\",\"type\":\"Legend\"},{\"attributes\":{},\"id\":\"3495\",\"type\":\"ResetTool\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"3505\"},\"glyph\":{\"id\":\"3506\"},\"group\":null,\"hover_glyph\":{\"id\":\"3508\"},\"muted_glyph\":{\"id\":\"3509\"},\"nonselection_glyph\":{\"id\":\"3527\"},\"selection_glyph\":{\"id\":\"3526\"},\"view\":{\"id\":\"3511\"}},\"id\":\"3510\",\"type\":\"GlyphRenderer\"},{\"attributes\":{\"tools\":[{\"id\":\"3491\"},{\"id\":\"3492\"},{\"id\":\"3493\"},{\"id\":\"3494\"},{\"id\":\"3495\"},{\"id\":\"3496\"},{\"id\":\"3524\"},{\"id\":\"3528\"}]},\"id\":\"3498\",\"type\":\"Toolbar\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.8},\"fill_color\":{\"value\":\"red\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"3526\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"3496\",\"type\":\"HelpTool\"},{\"attributes\":{\"mode\":\"append\",\"overlay\":{\"id\":\"3548\"},\"renderers\":[{\"id\":\"3510\"}]},\"id\":\"3528\",\"type\":\"BoxSelectTool\"},{\"attributes\":{},\"id\":\"3492\",\"type\":\"WheelZoomTool\"},{\"attributes\":{\"axis\":{\"id\":\"3487\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"3490\",\"type\":\"Grid\"},{\"attributes\":{},\"id\":\"3488\",\"type\":\"BasicTicker\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"3530\"},\"glyph\":{\"id\":\"3531\"},\"group\":null,\"hover_glyph\":null,\"muted_glyph\":{\"id\":\"3533\"},\"nonselection_glyph\":{\"id\":\"3532\"},\"view\":{\"id\":\"3535\"}},\"id\":\"3534\",\"type\":\"GlyphRenderer\"},{\"attributes\":{\"callback\":null,\"mode\":\"vline\",\"renderers\":[{\"id\":\"3510\"}],\"tooltips\":null},\"id\":\"3524\",\"type\":\"HoverTool\"},{\"attributes\":{},\"id\":\"3484\",\"type\":\"BasicTicker\"},{\"attributes\":{\"data\":{\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float32\",\"order\":\"little\",\"shape\":[100]},\"y\":{\"__ndarray__\":\"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\",\"dtype\":\"float32\",\"order\":\"little\",\"shape\":[100]}},\"selected\":{\"id\":\"3520\"},\"selection_policy\":{\"id\":\"3519\"}},\"id\":\"3505\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"red\"},\"hatch_alpha\":{\"value\":0.5},\"line_alpha\":{\"value\":0.5},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3508\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"3479\",\"type\":\"LinearScale\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.5},\"fill_color\":{\"value\":\"midnightblue\"},\"line_color\":{\"value\":\"white\"}},\"id\":\"3527\",\"type\":\"Circle\"},{\"attributes\":{},\"id\":\"3546\",\"type\":\"Selection\"}],\"root_ids\":[\"3550\",\"3618\"]},\"title\":\"Bokeh Application\",\"version\":\"2.4.3\"}};\n",
       "    var render_items = [{\"docid\":\"3947ceaa-53d1-484c-b22b-cd9e2db43368\",\"root_ids\":[\"3550\"],\"roots\":{\"3550\":\"5bb4e4b9-9773-4ba6-ae64-1832b032611c\"}}];\n",
       "    root.Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "    for (const render_item of render_items) {\n",
       "      for (const root_id of render_item.root_ids) {\n",
       "\tconst id_el = document.getElementById(root_id)\n",
       "\tif (id_el.children.length && (id_el.children[0].className === 'bk-root')) {\n",
       "\t  const root_el = id_el.children[0]\n",
       "\t  root_el.id = root_el.id + '-rendered'\n",
       "\t}\n",
       "      }\n",
       "    }\n",
       "  }\n",
       "  if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "        clearInterval(timer);\n",
       "        embed_document(root);\n",
       "      } else if (document.readyState == \"complete\") {\n",
       "        attempts++;\n",
       "        if (attempts > 200) {\n",
       "          clearInterval(timer);\n",
       "          console.log(\"Bokeh: ERROR: Unable to run BokehJS code because BokehJS library is missing\");\n",
       "        }\n",
       "      }\n",
       "    }, 25, root)\n",
       "  }\n",
       "})(window);</script>"
      ],
      "text/plain": [
       "Row\n",
       "    [0] HSpacer()\n",
       "    [1] Bokeh(Figure)\n",
       "    [2] HSpacer()"
      ]
     },
     "execution_count": 30,
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {
       "id": "3550"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p = scatter(np.array(x), np.array(y), title=\"Linear regression\")\n",
    "p.line(np.array(x), np.array(w_1 * x + w_0), line_width=6, line_alpha=0.6, color=\"red\", legend_label=r\"Learned model\")\n",
    "plot = (center_plot(p))\n",
    "plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "from cml.tasks import RegressionLinearSolver\n",
    "explorer = RegressionLinearSolver()\n",
    "global key\n",
    "key = random.PRNGKey(23)\n",
    "def solve(x, y):\n",
    "    global key\n",
    "    key_gd, key = random.split(key, 2)\n",
    "    w_1, w_0, loss_history = gradient_descent(key_gd, x, y)\n",
    "    x_predict = np.linspace(np.min(x), np.max(x), 200)\n",
    "    y_model = w_1 * x_predict + w_0\n",
    "    return np.array(x_predict), np.array(y_model) #np.array(jnp.zeros(y_model.shape))\n",
    "\n",
    "explorer.solve = solve\n",
    "explorer.render()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "TypeError: Gradient only defined for scalar-output functions. Output had shape: (1,).<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #192841; border-color: #779ecb\">\n",
    "\n",
    "> ### Going further\n",
    "\n",
    "> You are encouraged to experiment in a similar way as outlined for NumPy with different learning rates, maximum number of iterations, and regularization techniques, as well as compare the performance (both accuracy and speed) of the implemented algorithms with `scikit-learn` built-in linear regression functions.\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"capacity\"></a>\n",
    "## Understanding model capacity and selection\n",
    "\n",
    "\n",
    "In real-life problem, we are aiming to find the parameters of a model, but we do not really know what is the _real_ function underlying this process. So what we can decide to select _any_ function of _any_ **capacity** (complexity of the function). One of the problem with that, is that if we have a too simple function, it will _underfit_ (it is not complex enough for our observations). On the opposite end, if we have a function which is too complex, it might be able to _fit through all training points exactly_ ... even though there is noise in our observations ! This is examplified in the following\n",
    "\n",
    "<img src=\"images/01_soa_function_families.png\" align=\"center\"/>\n",
    "\n",
    "We can observe this idea and play with it directly by trying to find a function approximating our previous observations with a polynomial function chosen to have a degree inside \\([1,2,8]\\).\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Depending on the _capacity_ of the model, what we can observe is that\n",
    "\n",
    "- `capacity too low   -> underfitting   : prediction variance >  noise variance`\n",
    "- `adequate capacity  -> good fit       : prediction variance == noise variance`\n",
    "- `capacity too high  -> overfitting    : prediction variance <  noise variance`\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A similar example can be given for a classification problem in two dimensions as follows\n",
    "\n",
    "<img src=\"images/01_underfit.png\" align=\"center\"/>\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Exercises\n",
    "\n",
    "In the following, we define the exercises that you should fill for this session. These go further than what we have seen together in the course, but they are based on the exact same principles, simply with slightly more complex definitions. We provide an overall guideline for successfully implementing each of the exercise."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exercise 1 - Linear classification"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #013220; border-color: #03C03C\">\n",
    "\n",
    "> In this exercise, we will implement linear classification using NumPy and JAX. The goal of this exercise is to gain a better understanding of how linear classification works and how to implement it ourselves using our two first libraries of choice, namely NumPy and JAX. To complete this exercise, you will need to have a basic understanding of NumPy and JAX. You can use the resources provided in the previous exercises to learn more about these libraries.\n",
    "\n",
    "We help you out by first defining a simple linear classification problem to solve\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn.datasets import make_blobs\n",
    "# Properties of the problem\n",
    "n_observations = 200\n",
    "noise = 0.2\n",
    "c1_center = [-2, -1]\n",
    "c2_center = [2, 1]\n",
    "# Create points\n",
    "x_coords, y_class = make_blobs(n_samples=n_observations, centers=[c1_center, c2_center], n_features=2, cluster_std=0.55)\n",
    "x_data = x_coords + (noise * np.random.randn(n_observations, 2))\n",
    "#x_data = x_coords\n",
    "y_classes = y_class\n",
    "y_classes[y_classes == 0] = -1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #013220; border-color: #03C03C\">\n",
    "\n",
    "> ### Question 1.1 - Implement linear classification with NumPy\n",
    "\n",
    "> 1. Check the training data with two classes that are linearly separable.\n",
    "> 2. Initialize weights and bias.\n",
    "> 3. Implement the forward pass of the linear classifier.\n",
    "> 4. Implement the hinge loss for classification.\n",
    "> 5. Derive the gradients (backward pass) of the linear classifier.\n",
    "> 6. Implement gradient descent to optimize the weights and bias.\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {},
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.holoviews_exec.v0+json": "",
      "text/html": [
       "<div id='3962'>\n",
       "  <div class=\"bk-root\" id=\"429a05eb-f981-460f-83eb-31031550eba3\" data-root-id=\"3962\"></div>\n",
       "</div>\n",
       "<script type=\"application/javascript\">(function(root) {\n",
       "  function embed_document(root) {\n",
       "    var docs_json = {\"d5c55801-4cbf-4cb3-beb9-efb82016dcf2\":{\"defs\":[{\"extends\":null,\"module\":null,\"name\":\"ReactiveHTML1\",\"overrides\":[],\"properties\":[]},{\"extends\":null,\"module\":null,\"name\":\"FlexBox1\",\"overrides\":[],\"properties\":[{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_content\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"align_items\"},{\"default\":\"row\",\"kind\":null,\"name\":\"flex_direction\"},{\"default\":\"wrap\",\"kind\":null,\"name\":\"flex_wrap\"},{\"default\":\"flex-start\",\"kind\":null,\"name\":\"justify_content\"}]},{\"extends\":null,\"module\":null,\"name\":\"GridStack1\",\"overrides\":[],\"properties\":[{\"default\":\"warn\",\"kind\":null,\"name\":\"mode\"},{\"default\":null,\"kind\":null,\"name\":\"ncols\"},{\"default\":null,\"kind\":null,\"name\":\"nrows\"},{\"default\":true,\"kind\":null,\"name\":\"allow_resize\"},{\"default\":true,\"kind\":null,\"name\":\"allow_drag\"},{\"default\":[],\"kind\":null,\"name\":\"state\"}]},{\"extends\":null,\"module\":null,\"name\":\"click1\",\"overrides\":[],\"properties\":[{\"default\":\"\",\"kind\":null,\"name\":\"terminal_output\"},{\"default\":\"\",\"kind\":null,\"name\":\"debug_name\"},{\"default\":0,\"kind\":null,\"name\":\"clears\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationAreaBase1\",\"overrides\":[],\"properties\":[{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"}]},{\"extends\":null,\"module\":null,\"name\":\"NotificationArea1\",\"overrides\":[],\"properties\":[{\"default\":[],\"kind\":null,\"name\":\"notifications\"},{\"default\":\"bottom-right\",\"kind\":null,\"name\":\"position\"},{\"default\":0,\"kind\":null,\"name\":\"_clear\"},{\"default\":[{\"background\":\"#ffc107\",\"icon\":{\"className\":\"fas fa-exclamation-triangle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"warning\"},{\"background\":\"#007bff\",\"icon\":{\"className\":\"fas fa-info-circle\",\"color\":\"white\",\"tagName\":\"i\"},\"type\":\"info\"}],\"kind\":null,\"name\":\"types\"}]},{\"extends\":null,\"module\":null,\"name\":\"Notification\",\"overrides\":[],\"properties\":[{\"default\":null,\"kind\":null,\"name\":\"background\"},{\"default\":3000,\"kind\":null,\"name\":\"duration\"},{\"default\":null,\"kind\":null,\"name\":\"icon\"},{\"default\":\"\",\"kind\":null,\"name\":\"message\"},{\"default\":null,\"kind\":null,\"name\":\"notification_type\"},{\"default\":false,\"kind\":null,\"name\":\"_destroyed\"}]},{\"extends\":null,\"module\":null,\"name\":\"TemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]},{\"extends\":null,\"module\":null,\"name\":\"MaterialTemplateActions1\",\"overrides\":[],\"properties\":[{\"default\":0,\"kind\":null,\"name\":\"open_modal\"},{\"default\":0,\"kind\":null,\"name\":\"close_modal\"}]}],\"roots\":{\"references\":[{\"attributes\":{\"axis\":{\"id\":\"3925\"},\"coordinates\":null,\"dimension\":1,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"3928\",\"type\":\"Grid\"},{\"attributes\":{\"coordinates\":null,\"data_source\":{\"id\":\"3943\"},\"glyph\":{\"id\":\"3945\"},\"group\":null,\"hover_glyph\":null,\"muted_glyph\":{\"id\":\"3947\"},\"nonselection_glyph\":{\"id\":\"3946\"},\"view\":{\"id\":\"3949\"}},\"id\":\"3948\",\"type\":\"GlyphRenderer\"},{\"attributes\":{\"background_fill_alpha\":0.2,\"background_fill_color\":\"darkgrey\",\"border_line_alpha\":0.8,\"border_line_color\":\"grey\",\"border_line_width\":3,\"coordinates\":null,\"group\":null,\"items\":[{\"id\":\"3961\"}],\"label_text_color\":\"white\"},\"id\":\"3960\",\"type\":\"Legend\"},{\"attributes\":{},\"id\":\"3957\",\"type\":\"UnionRenderers\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.2},\"fill_color\":{\"field\":\"color\"},\"hatch_alpha\":{\"value\":0.2},\"hatch_color\":{\"field\":\"color\"},\"line_alpha\":{\"value\":0.2},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3947\",\"type\":\"Circle\"},{\"attributes\":{\"source\":{\"id\":\"3943\"}},\"id\":\"3949\",\"type\":\"CDSView\"},{\"attributes\":{},\"id\":\"3958\",\"type\":\"Selection\"},{\"attributes\":{},\"id\":\"3932\",\"type\":\"SaveTool\"},{\"attributes\":{\"label\":{\"field\":\"label\"},\"renderers\":[{\"id\":\"3948\"}]},\"id\":\"3961\",\"type\":\"LegendItem\"},{\"attributes\":{\"axis_label\":\"x\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"3955\"},\"group\":null,\"major_label_policy\":{\"id\":\"3956\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"3922\"}},\"id\":\"3921\",\"type\":\"LinearAxis\"},{\"attributes\":{\"bottom_units\":\"screen\",\"coordinates\":null,\"fill_alpha\":0.5,\"fill_color\":\"lightgrey\",\"group\":null,\"left_units\":\"screen\",\"level\":\"overlay\",\"line_alpha\":1.0,\"line_color\":\"black\",\"line_dash\":[4,4],\"line_width\":2,\"right_units\":\"screen\",\"syncable\":false,\"top_units\":\"screen\"},\"id\":\"3935\",\"type\":\"BoxAnnotation\"},{\"attributes\":{},\"id\":\"3926\",\"type\":\"BasicTicker\"},{\"attributes\":{\"axis_label\":\"y\",\"axis_label_standoff\":10,\"axis_label_text_color\":\"#FFFFFF\",\"axis_label_text_font\":\"Josefin Sans\",\"axis_label_text_font_size\":\"1.9em\",\"axis_label_text_font_style\":\"bold\",\"axis_line_alpha\":0.4,\"axis_line_color\":\"#E0E0E0\",\"coordinates\":null,\"formatter\":{\"id\":\"3952\"},\"group\":null,\"major_label_policy\":{\"id\":\"3953\"},\"major_label_text_color\":\"#E0E0E0\",\"major_label_text_font\":\"Josefin Sans\",\"major_label_text_font_size\":\"1.15em\",\"major_tick_line_alpha\":0.4,\"major_tick_line_color\":\"#E0E0E0\",\"minor_tick_line_alpha\":0.4,\"minor_tick_line_color\":\"#E0E0E0\",\"ticker\":{\"id\":\"3926\"}},\"id\":\"3925\",\"type\":\"LinearAxis\"},{\"attributes\":{},\"id\":\"3915\",\"type\":\"DataRange1d\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.1},\"fill_color\":{\"field\":\"color\"},\"hatch_alpha\":{\"value\":0.1},\"hatch_color\":{\"field\":\"color\"},\"line_alpha\":{\"value\":0.1},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3946\",\"type\":\"Circle\"},{\"attributes\":{\"fill_alpha\":{\"value\":0.6},\"fill_color\":{\"field\":\"color\"},\"hatch_alpha\":{\"value\":0.6},\"hatch_color\":{\"field\":\"color\"},\"line_alpha\":{\"value\":0.6},\"line_color\":{\"value\":\"white\"},\"size\":{\"value\":10},\"x\":{\"field\":\"x\"},\"y\":{\"field\":\"y\"}},\"id\":\"3945\",\"type\":\"Circle\"},{\"attributes\":{\"background_fill_alpha\":0.9,\"background_fill_color\":\"#0f0f0f\",\"below\":[{\"id\":\"3921\"}],\"border_fill_alpha\":0.8,\"border_fill_color\":\"#111111\",\"center\":[{\"id\":\"3924\"},{\"id\":\"3928\"},{\"id\":\"3960\"}],\"height\":450,\"left\":[{\"id\":\"3925\"}],\"outline_line_alpha\":0.25,\"outline_line_color\":\"#E0E0E0\",\"outline_line_width\":3,\"renderers\":[{\"id\":\"3948\"}],\"right\":[{\"id\":\"3960\"}],\"title\":{\"id\":\"3911\"},\"toolbar\":{\"id\":\"3936\"},\"width\":700,\"x_range\":{\"id\":\"3913\"},\"x_scale\":{\"id\":\"3917\"},\"y_range\":{\"id\":\"3915\"},\"y_scale\":{\"id\":\"3919\"}},\"id\":\"3910\",\"subtype\":\"Figure\",\"type\":\"Plot\"},{\"attributes\":{},\"id\":\"3917\",\"type\":\"LinearScale\"},{\"attributes\":{\"tools\":[{\"id\":\"3929\"},{\"id\":\"3930\"},{\"id\":\"3931\"},{\"id\":\"3932\"},{\"id\":\"3933\"},{\"id\":\"3934\"}]},\"id\":\"3936\",\"type\":\"Toolbar\"},{\"attributes\":{\"children\":[{\"id\":\"3963\"},{\"id\":\"3910\"},{\"id\":\"3984\"}],\"margin\":[0,0,0,0],\"name\":\"Row00293\"},\"id\":\"3962\",\"type\":\"Row\"},{\"attributes\":{},\"id\":\"3913\",\"type\":\"DataRange1d\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00290\",\"sizing_mode\":\"stretch_width\"},\"id\":\"3984\",\"type\":\"Spacer\"},{\"attributes\":{},\"id\":\"3929\",\"type\":\"PanTool\"},{\"attributes\":{\"margin\":[5,5,5,5],\"name\":\"HSpacer00289\",\"sizing_mode\":\"stretch_width\"},\"id\":\"3963\",\"type\":\"Spacer\"},{\"attributes\":{},\"id\":\"3934\",\"type\":\"HelpTool\"},{\"attributes\":{\"overlay\":{\"id\":\"3935\"}},\"id\":\"3931\",\"type\":\"BoxZoomTool\"},{\"attributes\":{},\"id\":\"3953\",\"type\":\"AllLabels\"},{\"attributes\":{\"coordinates\":null,\"group\":null,\"text\":\"Classification\",\"text_color\":\"#E0E0E0\",\"text_font\":\"Josefin Sans\",\"text_font_size\":\"2.1em\"},\"id\":\"3911\",\"type\":\"Title\"},{\"attributes\":{},\"id\":\"3922\",\"type\":\"BasicTicker\"},{\"attributes\":{},\"id\":\"3930\",\"type\":\"WheelZoomTool\"},{\"attributes\":{},\"id\":\"3955\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{},\"id\":\"3933\",\"type\":\"ResetTool\"},{\"attributes\":{},\"id\":\"3919\",\"type\":\"LinearScale\"},{\"attributes\":{},\"id\":\"3956\",\"type\":\"AllLabels\"},{\"attributes\":{},\"id\":\"3952\",\"type\":\"BasicTickFormatter\"},{\"attributes\":{\"axis\":{\"id\":\"3921\"},\"coordinates\":null,\"grid_line_alpha\":0.15,\"grid_line_color\":\"#E0E0E0\",\"group\":null,\"ticker\":null},\"id\":\"3924\",\"type\":\"Grid\"},{\"attributes\":{\"data\":{\"color\":[\"firebrick\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"firebrick\",\"firebrick\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"chocolate\",\"firebrick\",\"firebrick\",\"chocolate\",\"chocolate\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"firebrick\",\"firebrick\",\"firebrick\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"firebrick\",\"firebrick\",\"firebrick\",\"firebrick\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"firebrick\",\"chocolate\",\"chocolate\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"firebrick\",\"firebrick\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"chocolate\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"chocolate\",\"chocolate\",\"firebrick\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"firebrick\",\"firebrick\",\"chocolate\",\"chocolate\",\"chocolate\",\"chocolate\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"firebrick\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"firebrick\",\"firebrick\",\"chocolate\",\"chocolate\",\"chocolate\",\"chocolate\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"chocolate\",\"firebrick\",\"chocolate\",\"firebrick\",\"chocolate\"],\"label\":[\"class 2\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 2\",\"class 2\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 5\",\"class 2\",\"class 2\",\"class 5\",\"class 5\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 2\",\"class 2\",\"class 2\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 2\",\"class 2\",\"class 2\",\"class 2\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 2\",\"class 5\",\"class 5\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 2\",\"class 2\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 5\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 5\",\"class 5\",\"class 2\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 2\",\"class 2\",\"class 5\",\"class 5\",\"class 5\",\"class 5\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 2\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 2\",\"class 2\",\"class 5\",\"class 5\",\"class 5\",\"class 5\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 5\",\"class 2\",\"class 5\",\"class 2\",\"class 5\"],\"x\":{\"__ndarray__\":\"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\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[200]},\"y\":{\"__ndarray__\":\"OTILfrLt6D9++Wl2shflP7SnWYqpNu0/cE0VSmZb57+cgHoid4HeP/qU47z5QvA/WN4WYZXz8T8YqRGOBM7kP86VMCR3Y+O/aFW7iTN+6T8xeDZH433AP9ra3aU3Ufc/EShdG+ji/r91WiTLVCTrP9350UDmNdQ/+3QO9Cio6L/7ZMgqyuzwP1xcs1gh+s8/aZZgVDCs7j+B6OR1xI7wP7wn5Qoz6fM/Q7U05mB68z9tbhCY7Erzv5S+n6TMYN8/nJMfJHmD9L8shrRijVHjvw8znNri6vY/x24D3dboAMB03eEysMv2P8gCr+W+7Jm/oy2rPv96zL+OkZBiNFv3P2Mq2chpMPS/iwSyroRl+L9W0arrMZL9v3GV9KERYeI/wssVJWuh4L863wOGJs3LPySt1DBow/W/BcUJN5sy8D/ygCo6/4L3vywel+Mowe4/PF7p4EU47D9mfL+Kka3XP9DR/w5MQbk/gZH03uT86T+Q8DGbVFXtv+pXDhKAAPS/7EdHSqUw278Zg0XXOtjuP1WjfKecw+k/iE3BOgqv5L9CA/NBHUn0v5UFdR0s+PK/omsGehDNyb/MV6St6F70P/iCKDgTr+G/4Y3RD/OU+78znk390kbyP7rwOEQy2O8/KL30QqEA8T9YBH9l3QbpP5onefxLpuA/4b5trCa2/L9azvuE4h79P0bQSIpRlee/V2br9kbZ/L8qDclqDLj1PymVUm6MsOy/bVsHcwn6/7/Z2h6m97z7P3hI32958u2/p01lINEz978REx7VmDDzv9xSvwLPbds/lK6K9ZzXtD9oxq2W0N7kv8puTT9SCNU//FAh1U/tvL9j6WqD4XnRP3ZwQRbXdvQ//8n9EfDv9j9yHk1Iw3D5P7xDoUuH+O0/w4ZQ+7zD2T9duOj/LjnKPycYx9VfL9u/26LkEl2Q4T/EeSjzIUjsP3jBXM5uROQ/dbGTUkJq7b/X5k6cCvD4v6/CxCp20Oa/QOANFvTXyL/GWmoKsvfwP3opiKN6OuC/GJlAFPY36L+DDGw8w538PzMvYbMHZ/K/4gR0vLB36z9sbXa78KT6P4H0Q3K/UvG/ZGtb5h5GvL+n7qN9RMT7v7W/ecRYfey/wMocUhMw0D81k+6dskjwP+gAGDYyJ/M/gVd9ab4i7D9VazV2K0Lmv+uumcoycvE/m76TppB02r8AS3oE0HflP/Q8U/1lDO+/VuAfOe770j/vxyoWbhb+P5cz/CgYd/2/Qp0vb2/S678bVl0tDM3rv2BZqO8AEuo/fYdS+d5UzT/+GXJi4FLoP9AjJg8YjNo/ionRCAKq979QFZHrq9fpv/pI26NSBPU/ikbyFXv+4r8agzhH3tv3v7HxP9uQ0ea/ypa/J6j6+z+XA+ijxlkAwE5rtQIrffU/iXhGv2vm+z/+jQE9rMD1vzwxalmHRPQ/uzqFy2zc5T+dbEFDPyPvv8YA/ni4N+4/QOCAjR+Y8j+oX4xZWNr0v/avofuT7O8/KnF439vn17+CfjWXISv4v8Yh3sV3YwHARg89fsSg77/nLzmsTajBP2KRPWb3eeo/ELOIZbJstb+aDvMZ75LtvwNkuBJDMgBAqH3Q5TJu378tbjt7Zln/P8zQzh1rVPs/V6YrXHGQ8z/YEdqY8+frP0PYsVFFZ/G/PazDevVY6L/Uu55k+aXgv7yYRHmquP2/AH+Bv6e6PD8x2pREEub3vwR/IrP1Xe0/clZv3i3q/79FB0ReaOEMQFeoFQyLT++/PpG0I2uN0L9QDp9BZ/v4P4vpOLH4FuQ/O0xFrn+K8j8q/2s7y1rnv5wGWGe+pbo/DLHx5JfUAEAGpRMrZ4b7vy5WwPWEOPI/X0jdGfGZ478ILS9qL3/uvyQKP+slZvm//l5S0XFO+T8SBq+LyQTlv6pSEcLkTfs/uS+27JFE9T/1FCd6swL9P81oYlayytu/jOlthg7F/L++9St9Jmr3v0YnRioFCe6/BCY6XmlstD8j/hWURV/Uv0Abdocvc/Y/2JgyAG629r/ssMSg2LnpPyNB65bktOC/Ljc/RGzp8b8SbGmE0cjQPy7a1No4Zsq/wAY6itmX6L+PxJO7O4b8Pxea8EBxvfW/8yTud7hG5T+nybCWtgjwvw==\",\"dtype\":\"float64\",\"order\":\"little\",\"shape\":[200]}},\"selected\":{\"id\":\"3958\"},\"selection_policy\":{\"id\":\"3957\"}},\"id\":\"3943\",\"type\":\"ColumnDataSource\"},{\"attributes\":{\"client_comm_id\":\"bfd32a0d981f416482a7d01e508f1cf3\",\"comm_id\":\"9019d06a1a664f40add05a0d7827aacb\",\"plot_id\":\"3962\"},\"id\":\"4015\",\"type\":\"panel.models.comm_manager.CommManager\"}],\"root_ids\":[\"3962\",\"4015\"]},\"title\":\"Bokeh Application\",\"version\":\"2.4.3\"}};\n",
       "    var render_items = [{\"docid\":\"d5c55801-4cbf-4cb3-beb9-efb82016dcf2\",\"root_ids\":[\"3962\"],\"roots\":{\"3962\":\"429a05eb-f981-460f-83eb-31031550eba3\"}}];\n",
       "    root.Bokeh.embed.embed_items_notebook(docs_json, render_items);\n",
       "    for (const render_item of render_items) {\n",
       "      for (const root_id of render_item.root_ids) {\n",
       "\tconst id_el = document.getElementById(root_id)\n",
       "\tif (id_el.children.length && (id_el.children[0].className === 'bk-root')) {\n",
       "\t  const root_el = id_el.children[0]\n",
       "\t  root_el.id = root_el.id + '-rendered'\n",
       "\t}\n",
       "      }\n",
       "    }\n",
       "  }\n",
       "  if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "    embed_document(root);\n",
       "  } else {\n",
       "    var attempts = 0;\n",
       "    var timer = setInterval(function(root) {\n",
       "      if (root.Bokeh !== undefined && root.Bokeh.Panel !== undefined) {\n",
       "        clearInterval(timer);\n",
       "        embed_document(root);\n",
       "      } else if (document.readyState == \"complete\") {\n",
       "        attempts++;\n",
       "        if (attempts > 200) {\n",
       "          clearInterval(timer);\n",
       "          console.log(\"Bokeh: ERROR: Unable to run BokehJS code because BokehJS library is missing\");\n",
       "        }\n",
       "      }\n",
       "    }, 25, root)\n",
       "  }\n",
       "})(window);</script>"
      ],
      "text/plain": [
       "Row\n",
       "    [0] HSpacer()\n",
       "    [1] Bokeh(Figure)\n",
       "    [2] HSpacer()"
      ]
     },
     "execution_count": 33,
     "metadata": {
      "application/vnd.holoviews_exec.v0+json": {
       "id": "3962"
      }
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#1.\n",
    "from cml.plot import scatter_classes\n",
    "\n",
    "p = scatter_classes(x_data, y_classes, title=\"Classification\")\n",
    "plot = (center_plot(p))\n",
    "plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "from cml.plot import scatter_classes\n",
    "import time\n",
    "\n",
    "#2.\n",
    "w_1 = np.random.randn()\n",
    "w_0 = np.random.randn()\n",
    "bias = np.random.randn()\n",
    "weights = np.array([w_0,w_1,bias])\n",
    "\n",
    "\n",
    "n_iter = 1000\n",
    "lr = 0.001\n",
    "\n",
    "#4.\n",
    "def hinge_loss(y, y_bar):\n",
    "    return max(0,1-y*y_bar)\n",
    "\n",
    "#3.\n",
    "y_bar = np.zeros((200,1))    \n",
    "\n",
    "#5. 6.\n",
    "np_start_time = time.time()\n",
    "\n",
    "for _ in range(n_iter):\n",
    "    \n",
    "    for i in range(x_data.shape[0]):\n",
    "        y_bar[i] = np.matmul(weights,np.array([x_data[i][0],x_data[i][1],1]))\n",
    "    \n",
    "    dw_0 = 0\n",
    "    dw_1 = 0\n",
    "    dw_bias =  0\n",
    "    for j in range(x_data.shape[0]):\n",
    "    # Calculate the gradients and add them to the already existing gradient change\n",
    "        if(y_classes[j]*y_bar[j] < 1):\n",
    "            dw_0 += -y_classes[j]*x_data[j][0]\n",
    "            dw_1 += -y_classes[j]*x_data[j][1]\n",
    "            dw_bias += -y_classes[j]*1 #we have assigned this label one at all points\n",
    "\n",
    "    # Update the parameters\n",
    "    w_0 -= lr * dw_0\n",
    "    w_1 -= lr * dw_1\n",
    "    bias -= lr * dw_bias\n",
    "    \n",
    "np_end_time = time.time()\n",
    "\n",
    "\n",
    "#sanity check\n",
    "np_error = 0\n",
    "plot_y_np = w_0 * x_data[:,0]  + w_1*x_data[:,1] + bias\n",
    "\n",
    "for i in range(x_data.shape[0]):\n",
    "    y_test = w_0 * x_data[i][0] + w_1*x_data[i][1] + bias\n",
    "    if y_test > 0:\n",
    "        y_test = 1\n",
    "    else:\n",
    "        y_test = -1\n",
    "    \n",
    "    if y_test != y_classes[i]:\n",
    "        np_error += 1\n",
    "        \n",
    "np_error = (np_error/x_data.shape[0])*100\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #192841; border-color: #779ecb\">\n",
    "\n",
    "> ### Going further (optional)\n",
    "> Try to replace the Hinge loss by the cross-entropy loss in your implementation\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #013220; border-color: #03C03C\">\n",
    "\n",
    "> ### Question 1.2 - Implement linear classification with JAX\n",
    "\n",
    "> 1. Initialize weights and bias.\n",
    "> 2. Define the loss function using the jax.nn.softmax_cross_entropy_with_logits function.\n",
    "> 3. Implement gradient descent to optimize the weights and bias.\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1.3702862] [0.9594967] [0.2983443]\n"
     ]
    }
   ],
   "source": [
    "#1.\n",
    "import jax\n",
    "\n",
    "key = random.PRNGKey(42)\n",
    "key_0, key_1, key = random.split(key, 3)\n",
    "    \n",
    "w_1 = random.normal(key_1, (1,))\n",
    "w_0 = random.normal(key_0, (1,))\n",
    "bias = random.normal(key, (1,))\n",
    "\n",
    "\n",
    "n_iter = 1000\n",
    "lr = 0.001\n",
    "\n",
    "\n",
    "#2.\n",
    "def hinge_loss_jnp(x, y, w_0, w_1, bias):\n",
    "    y_pred = w_0 * x[0] + w_1 * x[1] + bias * x[2]\n",
    "    return jnp.maximum(0,1-(y*y_pred)).sum()\n",
    "\n",
    "grad_loss_function = jit(grad(hinge_loss_jnp, argnums=[2, 3, 4]))\n",
    "\n",
    "#3.\n",
    "jax_start_time = time.time()\n",
    "\n",
    "for _ in range(n_iter):\n",
    "    \n",
    "    one = jnp.ones([x_data.shape[0],1])\n",
    "    x = jnp.concatenate((x_data,one),axis=1)\n",
    "    \n",
    "    dw0 = 0\n",
    "    dw1 = 0\n",
    "    dw_bias =  0\n",
    "    \n",
    "    for i in range(x.shape[0]): #only got it to work elementwise, so we iterate over all elements to get a total hinge loss\n",
    "        gradients = grad_loss_function(x[i], y_classes[i], w_0, w_1, bias)\n",
    "        dw0 += gradients[0]\n",
    "        dw1 += gradients[1]\n",
    "        dw_bias += gradients[2]\n",
    "            \n",
    "    w_0 -= lr * dw0\n",
    "    w_1 -= lr * dw1\n",
    "    bias -= lr* dw_bias\n",
    "    \n",
    "jax_end_time = time.time()\n",
    "\n",
    "#sanity check\n",
    "jax_error = 0\n",
    "plot_y_jax = w_0 * x_data[:,0]  + w_1*x_data[:,1] + bias\n",
    "\n",
    "for i in range(x_data.shape[0]):\n",
    "    y_test = w_0 * x_data[i][0] + w_1*x_data[i][1] + bias\n",
    "    if y_test > 0:\n",
    "        y_test = 1\n",
    "    else:\n",
    "        y_test = -1\n",
    "    \n",
    "    if y_test != y_classes[i]:\n",
    "        jax_error += 1\n",
    "        \n",
    "jax_error = (jax_error/x_data.shape[0])*100\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #013220; border-color: #03C03C\">\n",
    "\n",
    "> ### Question 1.3 - Compare the results of the NumPy and JAX implementations.\n",
    "> 1. Plot the decision boundary for each implementation.\n",
    "> 2. Compare the training time and accuracy of each implementation.\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The decision boundary can be clearly seen by the separation of the two sets of points. Implementation difference split for colour.\n",
      "Time for Numpy implementation: 1.29329  seconds\n",
      "Time for JAX implementation: 93.70172  seconds\n",
      "Percentage error of Numpy implementation: 0.0 %\n",
      "Percentage error of JAX implementation: 0.0 %\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot\n",
    "import matplotlib.pyplot as plt\n",
    "ax = plt.axes(projection='3d')\n",
    "\n",
    "# Data for a three-dimensional line\n",
    "x0 = x_data[:,0]\n",
    "x1 = x_data[:,1]\n",
    "y_np = plot_y_np\n",
    "# ax.plot3D(x0line, x1line, y_np, 'gray')\n",
    "\n",
    "ax.scatter3D(x0, x1, y_np, c=y_np, cmap='Greens');\n",
    "\n",
    "y_jax = plot_y_jax\n",
    "ax.scatter3D(x0, x1, y_jax, c=y_np, cmap='Blues');\n",
    "\n",
    "print(\"The decision boundary can be clearly seen by the separation of the two sets of points. Implementation difference split for colour.\")\n",
    "\n",
    "#training time\n",
    "print(\"Time for Numpy implementation:\", round(np_end_time - np_start_time,5),\" seconds\")\n",
    "print(\"Time for JAX implementation:\", round(jax_end_time - jax_start_time,5),\" seconds\")\n",
    "\n",
    "#accuracy\n",
    "print(\"Percentage error of Numpy implementation:\", np_error,\"%\")\n",
    "print(\"Percentage error of JAX implementation:\", jax_error,\"%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #192841; border-color: #779ecb\">\n",
    "\n",
    "> ### Going further (optional)\n",
    "> Implement a more complex dataset and compare the results of the NumPy and JAX implementations.\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exercise 2 - Polynomial regression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "71643353<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #013220; border-color: #03C03C\">\n",
    "\n",
    "> In this exercise, you will implement polynomial regression using NumPy and JAX. This should be a quite straightforward extension of what we have seen until now. We deliberately removed details of the implementation (list of points to adress) so that you can start defining your first entire machine learning setup by yourself.\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "degree = 10\n",
    "degree = degree + 1 #when you write degree 2 you get x^2 terms for example\n",
    "n_iter = 1000\n",
    "lr = 0.0015\n",
    "\n",
    "\n",
    "import time\n",
    "import jax\n",
    "\n",
    "#Generate own dataset first\n",
    "true_w0 = 2\n",
    "true_w1 = 3\n",
    "key_0, key_1, key = random.split(key, 3)\n",
    "x = random.uniform(key_0, (100,))\n",
    "y = true_w0 + true_w1 * x + (random.normal(key_1, (100,)) * 0.2)\n",
    "\n",
    "#Generate a full dataset first that contains points up to the n-th degree\n",
    "fullmat = np.zeros((100,degree))\n",
    "\n",
    "for i in range(100):\n",
    "    one_sample = np.zeros((degree))\n",
    "    for j in range(degree):\n",
    "        one_sample[j] = x[i]**(j)\n",
    "    fullmat[i] = one_sample\n",
    "\n",
    "# Splitting our dataset\n",
    "train_size = int(0.8 * len(x))\n",
    "x_train, x_valid = fullmat[:train_size][:], fullmat[train_size:][:]\n",
    "y_train, y_valid = y[:train_size], y[train_size:]    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #013220; border-color: #03C03C\">\n",
    "\n",
    "> ### Question 2.1 - Implement polynomial regression using NumPy with gradient descent.\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "absolute error 3.1675105\n"
     ]
    }
   ],
   "source": [
    "def gradient_descent(x, y, n_iter, lr, degree):\n",
    "    # Initialize the parameters\n",
    "    weights = np.random.random(degree)\n",
    "\n",
    "    # Perform gradient descent\n",
    "    for i in range(n_iter):\n",
    "        # Calculate the predictions\n",
    "        y_bar = np.matmul(weights.transpose(),x.transpose())\n",
    "        # Calculate the gradients\n",
    "        dw_w = np.zeros((degree,1))\n",
    "        for j in range(degree):\n",
    "            dw_w[j] = 2 * np.sum((y_bar - y) * (x[:,j]))\n",
    "                                 \n",
    "        # Update the parameters\n",
    "        for j in range(weights.shape[0]):\n",
    "            weights[j] -= lr * dw_w[j]\n",
    "            \n",
    "    return weights\n",
    "\n",
    "np_start_time = time.time()\n",
    "\n",
    "weights = gradient_descent(x_train, y_train, n_iter, lr, degree)\n",
    "\n",
    "np_end_time = time.time()\n",
    "\n",
    "#sanity check\n",
    "y_check = np.matmul(weights.transpose(),x_valid.transpose())\n",
    "\n",
    "np_error = np.sum(abs(y_check - y_valid))\n",
    "\n",
    "print (\"absolute error\",np_error)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #013220; border-color: #03C03C\">\n",
    "\n",
    "> ### Question 2.2. Implement polynomial regression using JAX with gradient descent.\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "absolute error 7.98759\n"
     ]
    }
   ],
   "source": [
    "key_0, key_1, key = random.split(key, 3)\n",
    "\n",
    "def loss_function(x, y, weights):\n",
    "    y_pred = jnp.matmul(weights.transpose(),x.transpose())\n",
    "    residuals = y_pred - y\n",
    "    return jnp.mean(residuals**2)\n",
    "\n",
    "grad_loss_function = jit(grad(loss_function, argnums=[2]))\n",
    "\n",
    "\n",
    "def gradient_descent(key, x, y, lr, n_iter,degree):\n",
    "    # m = x.shape\n",
    "    #Initialize random weights\n",
    "    weights = jnp.zeros((degree,1))\n",
    "    for i in range(weights.shape[0]):\n",
    "        key, key_x = random.split(key,2)\n",
    "        weights = weights.at[i].set(jax.random.uniform(key))\n",
    "    \n",
    "    for _ in range(n_iter):\n",
    "        gradients = grad_loss_function(x, y, weights)\n",
    "        for i in range(weights.shape[0]):\n",
    "            dw =  - (lr * gradients[0][i])\n",
    "            curr = weights[i]\n",
    "            weights = weights.at[i].set(curr+dw)\n",
    "    return weights\n",
    "\n",
    "\n",
    "key_gd, key = random.split(key, 2)\n",
    "\n",
    "jax_start_time = time.time()\n",
    "\n",
    "weights = gradient_descent(key_gd, x_train, y_train,lr, n_iter,degree)\n",
    "\n",
    "jax_end_time = time.time()\n",
    "\n",
    "#Sanity check\n",
    "y_check = np.matmul(weights.transpose(),x_valid.transpose())\n",
    "\n",
    "jax_error = np.sum(abs(y_check - y_valid))\n",
    "\n",
    "print (\"absolute error\",jax_error)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #013220; border-color: #03C03C\">\n",
    "\n",
    "> ### Question 2.3 - Compare the results of both implementations and the speed of the optimization algorithm.\n",
    "\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Error of Numpy implementation  3.1675105\n",
      "Error of JAX implementation  7.98759\n",
      "Time for Numpy implementation  0.6746  seconds\n",
      "Time for JAX implementation  18.91867  seconds\n"
     ]
    }
   ],
   "source": [
    "#results\n",
    "print(\"Error of Numpy implementation \", np_error)\n",
    "print(\"Error of JAX implementation \", jax_error)\n",
    "\n",
    "#speed of optimization\n",
    "\n",
    "print(\"Time for Numpy implementation \", round(np_end_time - np_start_time,5),\" seconds\")\n",
    "print(\"Time for JAX implementation \", round(jax_end_time - jax_start_time,5),\" seconds\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"audio\"></a>\n",
    "# Audio applications\n",
    "\n",
    "In order to test our algorithms on audio and music data, we will work with several datasets. We will both be using well-known state-of-art datasets (through the TF dataset system). But first, we will rely on a simple dataset (`Musclefish`) that should be downloaded on your local computer first from this [link](https://nubo.ircam.fr/index.php/s/ByK4QL7nE4Mq5MA)\n",
    "\n",
    "**The following set of instructions will perform that automatically for you :)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!curl https://nubo.ircam.fr/index.php/s/ByK4QL7nE4Mq5MA/download/musclefish.zip --output musclefish.zip\n",
    "!unzip musclefish.zip\n",
    "!mkdir data\n",
    "!mv musclefish data/"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dataset details\n",
    "\n",
    "  |**Type**|*Origin*|\n",
    "  |-------:|:---------|\n",
    "  |**Classification**|[*MuscleFish*](http://knight.cis.temple.edu/~vasilis/Courses/CIS750/Papers/muscle_fish.pdf) dataset|\n",
    "  |**Music-speech**|[*MIREX Recognition*](http://www.music-ir.org/mirex/wiki/2015:Music/Speech_Classification_and_Detection) set|\n",
    "  |**Source separation**|[*SMC Mirum*](http://smc.inesctec.pt/research/data-2/) dataset|\n",
    "  |**Speech recognition**|[*CMU Arctic*](http://festvox.org/cmu_arctic/) dataset|\n",
    "\n",
    "**Unzip the file and place the `musclefish` folder inside a `data` folder in the root of this notebook**\n",
    "For the first parts of the tutorial, we will mostly rely solely on the classification dataset. In order to facilitate the interactions, we provide a dataset class called `AudioSupervisedDataset` that will allow to import all audio datasets along the tutorials. This class also contains a set of holders and will allow us to perform batch-wise gradient descent.\n",
    "\n",
    "```Python\n",
    "class AudioSupervisedDataset():\n",
    "    \"\"\"\n",
    "    Helper class to import datasets\n",
    "    % class_path  : Path to the dataset (string)\n",
    "    % type       : Type of dataset (string: 'classify', 'plain', 'metadata')\n",
    "    \"\"\" \n",
    "    audio_files\n",
    "    labels\n",
    "    labels_names\n",
    "    num_examples \n",
    "    num_classes \n",
    "```\n",
    "\n",
    "We also provide a simplified function `import_dataset` that is demonstrated below.\n",
    "\n",
    "***\n",
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #192841; border-color: #779ecb\">\n",
    "\n",
    "**_Exercise_**  \n",
    "\n",
    "  1. Launch the import procedure and check the corresponding structure\n",
    "  2. Code a count function that prints the name and number of examples for each classes \n",
    "\n",
    "</div>\n",
    "\n",
    "***"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from cml.data import import_dataset, AudioSupervisedDataset\n",
    "data_path = \"data/\"\n",
    "batch_size = 16\n",
    "# Instantiate the dataset class\n",
    "train_dataset = import_dataset(data_path, \"musclefish\", 'train', batch_size=batch_size)\n",
    "# Generate a batch of data from the train dataset\n",
    "batch_x, batch_y = next(train_dataset)\n",
    "# Print the shape of the input and output batch tensors\n",
    "print(train_dataset.__len()__)\n",
    "print(batch_x.shape)  # (16, 44100)\n",
    "print(batch_y.shape)  # (16,)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#%% Q-0.1.2 - Count function to print the number of examples in each class along with class label\n",
    "import numpy as np\n",
    "n_batch = 0\n",
    "train_dataset._reset_generator()\n",
    "counter = np.zeros((16,1))\n",
    "for batch_x, batch_y in iter(train_dataset):\n",
    "    n_batch += 1\n",
    "    for elem in batch_y:\n",
    "        idx = elem.numpy()\n",
    "        counter[idx] += 1\n",
    "        \n",
    "print(counter)\n",
    "print(\"class label is just index position\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Preprocessing\n",
    "\n",
    "We will rely on a set of spectral transforms that allow to obtain a more descriptive view over the audio information. As most of these are out of the scope of the machine learning course, we redirect you to a [signal processing course](https://ccrma.stanford.edu/~jos/sasp/) proposed by [Julius O. Smith](https://ccrma.stanford.edu/~jos/).  \n",
    "\n",
    "The following functions to compute various types of transforms are given as part of the basic audio dataset class, in the `cml.data.audio` package\n",
    "\n",
    "  |**Name**|*Transform*|\n",
    "  |-------:|:----------|\n",
    "  |`stft`       |[Short-term Fourier transform](https://en.wikipedia.org/wiki/Short-time_Fourier_transform)|\n",
    "  |`mel`  |[Mel scale](https://en.wikipedia.org/wiki/Mel_scale) transform|\n",
    "  |`chroma` |[Chromas vector](https://en.wikipedia.org/wiki/Harmonic_pitch_class_profiles)|\n",
    "  |`cqt`        |[Constant-Q](https://en.wikipedia.org/wiki/Constant_Q_transform) transform|\n",
    "\n",
    "In order to perform the various computations, we provide the following function, in the `AudioSupervisedDataset` which performs the different transforms on a complete dataset.  \n",
    "\n",
    "``` Python\n",
    "dataset.transform(index, name)\n",
    "    \"\"\" index   : Specific index in our dataset \"\"\"\n",
    "    \"\"\" name    : Name of the transform to apply \"\"\"\n",
    "\n",
    "# The name can be selected in \n",
    "\"stft\"      # Power spectrum (STFT)\n",
    "\"mel\"       # Spectrum in Mel scale\n",
    "\"chroma\"    # Chroma vectors\n",
    "\"cqt\"       # Constant-Q transform\n",
    "```\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***\n",
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #192841; border-color: #779ecb\">\n",
    "\n",
    "**Exercise**  \n",
    "\n",
    "  1. Launch the transform computation procedure and check the corresponding structure\n",
    "  2. For each class, select a random element and plot its various transforms on a single plot. You should obtain plots similar to those shown afterwards.\n",
    "  3. For each transform, try to spot major pros and cons of their representation.\n",
    "  \n",
    "</div>\n",
    "\n",
    "***"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 0.2 - Pre-process the audio to obtain spectral transforms \n",
    "power_spec = train_dataset.transform(819, \"stft\")\n",
    "print(power_spec[511])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#%% Q-0.2.2 - Plot the various transforms \n",
    "\n",
    "######################\n",
    "# YOUR CODE GOES HERE\n",
    "######################\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Features\n",
    "\n",
    "<div markdown = \"1\">\n",
    "\n",
    "As you might have noted from the previous exercice, most spectral transforms have a very high dimensionality, and might not be suited to exhibit the relevant structure of different classes. To that end, we provide a set of functions for computing several spectral features in the `cml.data` package, we redirect interested readers to this [exhaustive article](http://recherche.ircam.fr/anasyn/peeters/ARTICLES/Peeters_2003_cuidadoaudiofeatures.pdf) on spectral features computation.\n",
    "\n",
    "  |**File**|*Transform*|\n",
    "  |-------:|:----------|\n",
    "  |`spectral_centroid`|Spectral centroid|\n",
    "  |`spectral_bandwidth`|Spectral bandwidth|\n",
    "  |`spectral_contrast`|Spectral contrast|\n",
    "  |`spectral_flatness`|Spectral flatness|\n",
    "  |`spectral_rolloff`|Spectral rolloff|\n",
    "\n",
    "Once again, we provide a function to perform the computation of different features on a complete set. Note that for each feature, we obtain the temporal evolution in a vector. Therefore, for further learning tasks, if you wish to obtain simplified spaces, you might need to compute the mean and standard deviation of each feature.\n",
    "\n",
    "``` Python\n",
    "dataset.feature(index, name)\n",
    "    \"\"\" index   : Specific index in our dataset \"\"\"\n",
    "    \"\"\" name    : Name of the feature to obtain \"\"\"\n",
    "\n",
    "# Name can be chosen inside the following\n",
    "\"loudness\"    # Loudness\n",
    "\"centroid\"    # Spectral centroid\n",
    "\"bandwidth\"   # Spectral bandwidth\n",
    "\"contrast\"    # Spectral contrast\n",
    "\"flatness\"    # Spectral flatness\n",
    "\"rolloff\"     # Spectral rolloff\n",
    "\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "***\n",
    "<div class=\"alert alert-success\" markdown=\"1\" style=\"color:white; background-color: #192841; border-color: #779ecb\">\n",
    "\n",
    "**Exercise**\n",
    "\n",
    "  1. Launch the feature computation procedure and check the corresponding structure\n",
    "  2. This time for each class, superimpose the plots of various features on a single plot, along with a boxplot of mean and standard deviations. You should obtain plots similar to those shown afterwards.\n",
    "  3. What conclusions can you make on the discriminative power of each feature ?\n",
    "  4. Perform scatter plots of the mean features for all the dataset, while coloring different classes.\n",
    "  5. What conclusions can you make on the discriminative power of mean features ?\n",
    "\n",
    "</div>\n",
    "\n",
    "***"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 0.3 - Compute temporal spectral features\n",
    "power_spec = train_dataset.feature(0, \"centroid\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#%% Q-0.3.2 - Plot the various features \n",
    "\n",
    "# Use these styles for boxplot\n",
    "boxprops=dict(linewidth=3, color='white')\n",
    "whiskerprops=dict(linewidth=3, color='white')\n",
    "medianprops=dict(linewidth=2.5, color='firebrick')\n",
    "flierprops = dict(markeredgecolor='white', markerfacecolor='firebrick')\n",
    "\n",
    "######################\n",
    "# YOUR CODE GOES HERE\n",
    "######################\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#%% Q-0.3.4 - Observe the distribution of classes for different features\n",
    "\n",
    "# This allows to use 3D rendering in matplotlib\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "plt.figure(figsize=(12,8))\n",
    "# Create a vector of random colors for each class\n",
    "colorVect = np.zeros((3, len(data_struct[\"class_names\"])));\n",
    "for c in range(len(data_struct[\"class_names\"])):\n",
    "    colorVect[:,c] = np.random.rand(3);\n",
    "\n",
    "######################\n",
    "# YOUR CODE GOES HERE\n",
    "######################\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "That's it for this tutorial, now remember that we can use any form of description (features) as a basis for learning algorithms. We will see in the next tutorial what we an do with these features."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.16"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autoclose": false,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}