Decision-Trees-For-Knowledge-Discovery-With-Scikit-Learn.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Discovering structure behind data\n",
    "\n",
    "Let's understand and model the hidden structure behind data with Decision Trees. In this tutorial, we'll explore and inspect how a model can do its decisions on a car evaluation data set. Decision trees work with simple \"if\" clauses dichotomically chained together, splitting the data flow recursively on those \"if\"s until they reach a leaf where we can categorize the data. Such data inspection could be used to reverse engineer the behavior of any function. \n",
    "\n",
    "Since [decision trees](http://www.r2d3.us/visual-intro-to-machine-learning-part-1/) are good algorithms for discovering the structure hidden behind data, we'll use and model the car evaluation data set, for which the prediction problem is a (deterministic) surjective function. This means that the inputs of the examples in the data set cover all the possibilities, and that for each possible input value, there is only one answer to predict (thus, two examples with the same input values would never have a different expected prediction). On the point of view of Data Science, because of the properties of our dataset, we won't need to use a test set nor to use cross validation. Thus, the error we will obtain below at modelizing our dataset would be equal to the true test error if we had a test set.\n",
    "\n",
    "The attribute to predict in the data set could have been, for example, created from a programmatic function and we will basically reverse engineer the logic mapping the inputs to the outputs to recreate the function and to be able to explain it visually.\n",
    "\n",
    "\n",
    "## About the Car Evaluation Data Set\n",
    "\n",
    "For more information: http://archive.ics.uci.edu/ml/datasets/Car+Evaluation\n",
    "\n",
    "### Overview\n",
    "\n",
    "The Car Evaluation Database was derived from a simple hierarchical decision model originally developed for the demonstration of DEX, M. Bohanec, V. Rajkovic: Expert system for decision making. Sistemica 1(1), pp. 145-157, 1990.). The model evaluates cars according to the following concept structure: \n",
    "\n",
    "- CAR car acceptability:\n",
    "  - PRICE overall price:\n",
    "    - **buying** buying price\n",
    "    - **maint** price of the maintenance\n",
    "  - TECH technical characteristics:\n",
    "    - COMFORT comfort:\n",
    "      - **doors** number of doors\n",
    "      - **persons** capacity in terms of persons to carry\n",
    "      - **lug_boot** the size of luggage boot\n",
    "    - **safety** estimated safety of the car\n",
    "\n",
    "Input attributes are printed in lowercase. Besides the target concept (CAR), the model includes three intermediate concepts: PRICE, TECH, COMFORT. Every concept is in the original model related to its lower level descendants by a set of examples. \n",
    "\n",
    "The Car Evaluation Database contains examples with the structural information removed, i.e., directly relates CAR to the six input attributes: buying, maint, doors, persons, lug_boot, safety. \n",
    "\n",
    "Because of known underlying concept structure, this database may be particularly useful for testing constructive induction and structure discovery methods. \n",
    "\n",
    "### Attributes, instances, and Class Distribution\n",
    "\n",
    "Number of Attributes: 6\n",
    "\n",
    "Missing Attribute Values: none\n",
    "\n",
    "| Attribute  | Values |\n",
    "|------------|--------|\n",
    "| buying     | v-high, high, med, low |\n",
    "| maint      | v-high, high, med, low |\n",
    "| doors      | 2, 3, 4, 5-more |\n",
    "| persons    | 2, 4, more |\n",
    "| lug_boot   | small, med, big |\n",
    "| safety     | low, med, high |\n",
    "\n",
    "Number of Instances: 1728 (Instances completely cover the attribute space.)\n",
    "\n",
    "| class | N | N[%] |\n",
    "|---|---|---|\n",
    "| unacc | 1210 | 70.023 % |\n",
    "| acc | 384 | 22.222 % |\n",
    "| good | 69 | 3.993 % |\n",
    "| v-good | 65 | 3.762 % |\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## We'll now load the car evaluation data set in Python and then train decision trees with Scikit-Learn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.preprocessing import LabelEncoder, OneHotEncoder\n",
    "from sklearn import tree\n",
    "import pydot\n",
    "\n",
    "from io import StringIO\n",
    "import os"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Define the features and preprocess the car evaluation data set\n",
    "\n",
    "We'll preprocess the attributes into redundant features, such as using an integer index (linear) to represent a value for an attribute, as well as also using a one-hot encoding for each attribute's possible values as new features. Despite the fact that this is redundant, this will help to make the tree smaller since it has more choice on how to split data on each branch. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Data set's shape,\n",
      "X.shape, integer_y.shape, len(feature_names), len(class_names):\n",
      "(1728, 27) (1728,) 27 4\n"
     ]
    }
   ],
   "source": [
    "# The integer values for features will take\n",
    "# a range from 0 to n-1 in the lists of possible values:\n",
    "input_labels = [\n",
    "    [\"buying\", [\"vhigh\", \"high\", \"med\", \"low\"]],\n",
    "    [\"maint\", [\"vhigh\", \"high\", \"med\", \"low\"]],\n",
    "    [\"doors\", [\"2\", \"3\", \"4\", \"5more\"]],  # Here indexes are not real values\n",
    "    [\"persons\", [\"2\", \"4\", \"more\"]],\n",
    "    [\"lug_boot\", [\"small\", \"med\", \"big\"]],\n",
    "    [\"safety\", [\"low\", \"med\", \"high\"]],\n",
    "]\n",
    "\n",
    "class_names = [\"unacc\", \"acc\", \"good\", \"vgood\"]\n",
    "\n",
    "# Load data set\n",
    "data = np.genfromtxt(os.path.join('data', 'car.data'), delimiter=',', dtype=\"U\")\n",
    "data_inputs = data[:, :-1]\n",
    "data_outputs = data[:, -1]\n",
    "\n",
    "def str_data_to_one_hot(data, input_labels):\n",
    "    \"\"\"Convert each feature's string to a flattened one-hot array. \"\"\"\n",
    "    X_int = LabelEncoder().fit_transform(data.ravel()).reshape(*data.shape)\n",
    "    X_bin = OneHotEncoder().fit_transform(X_int).toarray()\n",
    "    \n",
    "    feature_names = []\n",
    "    for a in input_labels:\n",
    "        key = a[0]\n",
    "        for b in a[1]:\n",
    "            value = b\n",
    "            feature_names.append(\"{}_is_{}\".format(key, value))\n",
    "\n",
    "    return X_bin, feature_names\n",
    "\n",
    "def str_data_to_linear(data, input_labels):\n",
    "    \"\"\"Convert each feature's string to an integer index\"\"\"\n",
    "    X_lin = np.array([[\n",
    "        input_labels[a][1].index(j) for a, j in enumerate(i)\n",
    "    ] for i in data])\n",
    "    \n",
    "    # Integer feature indexes will range\n",
    "    # from 0 to n-1 from indexes in the label list:\n",
    "    feature_names = [i[0] + \"_index\" for i in input_labels]\n",
    "    \n",
    "    return X_lin, feature_names\n",
    "\n",
    "# Take both one-hot and linear versions of input features: \n",
    "X_one_hot, feature_names_one_hot = str_data_to_one_hot(data_inputs, input_labels)\n",
    "X_linear_int, feature_names_linear_int = str_data_to_linear(data_inputs, input_labels)\n",
    "\n",
    "# Put that together:\n",
    "X = np.concatenate([X_one_hot, X_linear_int], axis=-1)\n",
    "feature_names = feature_names_one_hot + feature_names_linear_int\n",
    "\n",
    "# Outputs use indexes, this is not one-hot:\n",
    "integer_y = np.array([class_names.index(i) for i in data_outputs])\n",
    "\n",
    "print(\"Data set's shape,\")\n",
    "print(\"X.shape, integer_y.shape, len(feature_names), len(class_names):\")\n",
    "print(X.shape, integer_y.shape, len(feature_names), len(class_names))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train a simple decision tree to fit the data set:\n",
    "\n",
    "First, let's define some hyperparameters, such as the depth of the tree."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision tree trained!\n",
      "Errors: 6.53935185185 %\n",
      "Accuracy: 93.4606481481 %\n"
     ]
    }
   ],
   "source": [
    "max_depth = 6\n",
    "clf = tree.DecisionTreeClassifier(max_depth=max_depth)\n",
    "clf = clf.fit(X, integer_y)\n",
    "\n",
    "print(\"Decision tree trained!\")\n",
    "accuracy = clf.score(X, integer_y)\n",
    "print(\"Errors:\", 100 - accuracy * 100, \"%\")\n",
    "print(\"Accuracy:\", accuracy * 100, \"%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plot and save the tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "def plot_first_tree(clf, class_names, tree_name):\n",
    "    \"\"\"\n",
    "    Plot and save our scikit-learn tree.\n",
    "    \"\"\"\n",
    "    graph_save_path = os.path.join(\n",
    "        \"exported_sklearn_trees\", \n",
    "        \"{}\".format(tree_name)\n",
    "    )\n",
    "\n",
    "    tree.export_graphviz(clf, out_file=\"{}.dot\".format(graph_save_path))\n",
    "    dotfile = StringIO()\n",
    "    tree.export_graphviz(\n",
    "        clf, out_file=dotfile,\n",
    "        feature_names=feature_names, class_names=class_names,\n",
    "        filled=True, rotate=True\n",
    "    )\n",
    "    pydot.graph_from_dot_data(dotfile.getvalue())[0].write_png(\"{}.png\".format(graph_save_path))\n",
    "\n",
    "# Plot our simple tree:\n",
    "plot_first_tree(clf, class_names, tree_name=\"simple_tree\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that [the tree below can also be viewed here online](https://github.com/Vooban/Decision-Trees-For-Knowledge-Discovery/tree/master/exported_sklearn_trees).\n",
    "\n",
    "![simple tree](exported_sklearn_trees/simple_tree.png)\n",
    "\n",
    "### Plot the importance of each input features of the simple tree:\n",
    "\n",
    "Note here that it is the feature importance according to our simple, shallow tree. A fully complex trees would surely include more of the features/attributes, and with different proportions. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABA8AAAInCAYAAAD3b3C6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f610e2ddba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def feature_importance_chart(clf, classifier_name, feature_names):\n",
    "    sorted_feature_importances, sorted_feature_names = (\n",
    "        zip(*sorted(zip(clf.feature_importances_, feature_names)))\n",
    "    )\n",
    "    plt.figure(figsize=(16, 9))\n",
    "    plt.barh(range(len(sorted_feature_importances)), sorted_feature_importances)\n",
    "    plt.yticks(\n",
    "        range(len(sorted_feature_importances)),\n",
    "        [\"{}: {:.3}\".format(a, b) for a, b in zip(sorted_feature_names, sorted_feature_importances)]\n",
    "    )\n",
    "    plt.title(\"The Gini feature importance for the {} \\n\"\n",
    "              \"(total decrease in node impurity, weighted by the \"\n",
    "              \"probability of reaching that node)\".format(classifier_name))\n",
    "    plt.show()\n",
    "\n",
    "feature_importance_chart(clf, \"simple tree\", feature_names)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Let's now generate a fully perfect (complex) tree\n",
    "\n",
    "<p align=\"center\">\n",
    "  <a href=\"http://theinceptionbutton.com/\" ><img src=\"deeper.jpg\" /></a>\n",
    "</p>\n",
    "\n",
    "Let's [go deeper](http://theinceptionbutton.com/). Let's build a deeper tree. At least, a simple tree like the one above is interesting for having a simplfied view of the true logic behind our data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decision tree trained!\n",
      "Errors: 0.0 %\n",
      "Accuracy: 100.0 %\n"
     ]
    }
   ],
   "source": [
    "max_depth = None  # Full depth\n",
    "clf = tree.DecisionTreeClassifier(max_depth=max_depth)\n",
    "clf = clf.fit(X, integer_y)\n",
    "\n",
    "print(\"Decision tree trained!\")\n",
    "accuracy = clf.score(X, integer_y)\n",
    "print(\"Errors:\", 100 - accuracy * 100, \"%\")\n",
    "print(\"Accuracy:\", accuracy * 100, \"%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### A plot of the full tree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "plot_first_tree(clf, class_names, tree_name=\"complex_tree\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note that [the tree below can also be viewed here online](https://github.com/Vooban/Decision-Trees-For-Knowledge-Discovery/tree/master/exported_sklearn_trees). It would also be possible to [extract the tree as true code and create a function](https://stackoverflow.com/questions/20224526/how-to-extract-the-decision-rules-from-scikit-learn-decision-tree).\n",
    "\n",
    "![complex tree](exported_sklearn_trees/complex_tree.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Finally, the full feature importance:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABBwAAAInCAYAAAA79kpRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkjR0FhwkSZIkSdLQ/f9+v2jB9SkF6QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f6107f4ca58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "feature_importance_chart(clf, \"complex tree\", feature_names)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "## Conclusion\n",
    "\n",
    "To sum up, we managed to get good classification results and to be able to explain those results visually and automatically. Note that it would have been possible to solve a regression problem with the same algorithm, such as predicting a price rather than a category.\n",
    "\n",
    "Such a technique can be useful in reverse engineering an existing system, such as an old one that has been coded in a peculiar programming language and for which the employees who coded it have left. This technique can also be used for data mining, gaining business intelligence, and insights from data.\n",
    "\n",
    "In the case that your data does not represent a pure function like we have here, such as if for two of your input examples it is possible to have two possible different predictions, then a tree cannot model the data set with 100% accuracy. Hopefully, if you are in that situation where the logic behind the data is not perfect, it is possible to [repeat the experiment by using XGBoost](https://github.com/Vooban/Decision-Trees-For-Knowledge-Discovery/blob/master/Decision-Trees-For-Knowledge-Discovery-With-XGBoost.ipynb), which can help by incrementally training many trees to reduce the error and training them in an optimized fashion. The only disadvantage of that is that those boosted forests would be harder to explain due to the fact that you would have many trees."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[NbConvertApp] Converting notebook Decision-Trees-For-Knowledge-Discovery-With-Scikit-Learn.ipynb to markdown\n",
      "[NbConvertApp] Support files will be in Decision-Trees-For-Knowledge-Discovery-With-Scikit-Learn_files/\n",
      "[NbConvertApp] Making directory Decision-Trees-For-Knowledge-Discovery-With-Scikit-Learn_files\n",
      "[NbConvertApp] Making directory Decision-Trees-For-Knowledge-Discovery-With-Scikit-Learn_files\n",
      "[NbConvertApp] Writing 12276 bytes to Decision-Trees-For-Knowledge-Discovery-With-Scikit-Learn.md\n"
     ]
    }
   ],
   "source": [
    "# Let's convert this notebook to a README for the GitHub project's title page:\n",
    "!jupyter nbconvert --to markdown Decision-Trees-For-Knowledge-Discovery-With-Scikit-Learn.ipynb\n",
    "!mv Decision-Trees-For-Knowledge-Discovery-With-Scikit-Learn.md README.md"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}