training.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "f0ecd9b3-f92e-4ff5-ae50-650532c120e1",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "# Working with time series data with pandas 🐼"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e8fc83ef-f209-466c-95d3-38251f03f722",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "::::{grid}\n",
    "\n",
    ":::{card}\n",
    ":header: Context 🤔 \n",
    "Today we are building up our arsenal of data structures with a really exciting one - the `pandas` `Dataframe`.  This is a data structure that models **tabular data**, or data with labelled rows and columns. Learning this data structure will allow us to start manipulating large datasets of that format, and do things that we couldn't otherwise do in a program like excel.\n",
    ":::\n",
    "\n",
    ":::{card}\n",
    ":header: Outcomes 🎓 \n",
    "Create pandas dataframes, locate values within a dataframe, and run functions on a dataframe. \n",
    "\n",
    "Use pandas to open some example airborne data from the SARP data folder.\n",
    ":::\n",
    "\n",
    ":::{card}\n",
    ":header: Skills 🤓 \n",
    "Creating filepaths and navigating the data directory.\n",
    "::: \n",
    "::::"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ef40ad5b-4077-4f16-8b77-bab2f95759ba",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "## From Lists to DataFrames - Labelled Data\n",
    "\n",
    "Our first Python data structure was a List, which organized values in a certain order using an **index**. Indexing is common, but it requires the programmer to track the meaning of each item in the list in their head.\n",
    "\n",
    "As an example, let's take a list which represents several days of forecast Air Quality Index (AQI) data for Hampton Roads, VA.\n",
    "\n",
    "```{image} ./images/hampton_aqi.png\n",
    ":alt: Github logo\n",
    ":width: 400px\n",
    ":align: center\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "5404e43f-b7d2-4b74-9eb1-5521d5bdd85a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[38, 38, 69, 60, 61, 48, 39]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hampton_aqi_list = [38, 38, 69, 60, 61, 48, 39]\n",
    "\n",
    "hampton_aqi_list"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a1bc313-d7cb-4a3a-b541-e5fa96fc3d07",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "If someone wanted to know \"what is the AQI value for June 10th?\" they would have to figure out that Jun 10th is the 6th value in the list. It therefore has index 5 and is accessed like:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "fcc75b94-fb65-4762-b9dd-b7704cdfe2bc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "48"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hampton_aqi_list[5]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "47b8fdfa-bcb8-4457-9185-ede95a58b10d",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "Pandas is also a library for keeping track of a list of values, but pandas data is **labelled**. This means that in addition to indexes values can be extracted using names. Here is an example of a pandas dataframe, the core pandas data structure, with the same data in it:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "e54a1087-40cf-4fc4-8e94-65f41f2247a0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "482c0fb7-da75-4f9a-bb95-db6e7770d1c2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AQI</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2023-06-05</th>\n",
       "      <td>38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-06</th>\n",
       "      <td>38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-07</th>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-08</th>\n",
       "      <td>60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-09</th>\n",
       "      <td>61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-10</th>\n",
       "      <td>48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2023-06-11</th>\n",
       "      <td>39</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            AQI\n",
       "2023-06-05   38\n",
       "2023-06-06   38\n",
       "2023-06-07   69\n",
       "2023-06-08   60\n",
       "2023-06-09   61\n",
       "2023-06-10   48\n",
       "2023-06-11   39"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hampton_aqi_df = pd.DataFrame({'AQI': [38, 38, 69, 60, 61, 48, 39],}, \n",
    "                              index= ['2023-06-05', '2023-06-06', '2023-06-07', '2023-06-08', \n",
    "                                      '2023-06-09', '2023-06-10', '2023-06-11']\n",
    "                             )\n",
    "hampton_aqi_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "3183b372-86d2-46ab-a317-a2cedd2dfdc7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AQI    48\n",
       "Name: 2023-06-10, dtype: int64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hampton_aqi_df.loc['2023-06-10']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ccf2c9ea-de2e-4e53-9755-367314fd28d9",
   "metadata": {},
   "source": [
    "## 🗃️ Libraries"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "40685c77-4db5-4ddc-8cb0-ce209d42bfe8",
   "metadata": {},
   "source": [
    "Today we are going to dig into our first library in Python!  A **library** is a like an \"extras\" package for a programming language (think Settlers of Catan expansion pack).  It is a set of commands that doesn't automatically come activated when you use Python, but that you can still use if you just tell Python that you plan to do so.  You let python know you are going to use a library with an `import` statment.  After you do that the additional commands become availble."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "262d1779-e26e-4360-9d09-07319e79a710",
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'math' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Input \u001b[0;32mIn [1]\u001b[0m, in \u001b[0;36m<cell line: 2>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[38;5;66;03m# Attempting to use the `math` library before importing\u001b[39;00m\n\u001b[0;32m----> 2\u001b[0m \u001b[43mmath\u001b[49m\u001b[38;5;241m.\u001b[39mfactorial(\u001b[38;5;241m3\u001b[39m)\n",
      "\u001b[0;31mNameError\u001b[0m: name 'math' is not defined"
     ]
    }
   ],
   "source": [
    "# Attempting to use the `math` library before importing\n",
    "math.factorial(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "44dfeb63-9db1-4ab3-8e53-b8f0629705e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Importing a Python library\n",
    "import math"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "029a36f7-05af-4bb7-b8c6-74addd2de6bc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Attempt #2 to use the `math` libray, after importing\n",
    "math.factorial(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a4990be-3855-49d5-b437-a3af27fe8e5c",
   "metadata": {},
   "source": [
    "Once we have imported the library we can use the new commands that the library contains.  The `math` library was just an example and it isn't the focus of this lesson, but if you want you can see a list of `math` commands [here](https://docs.python.org/3/library/math.html)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "46bf579d-ee92-45a5-a90c-fc620508acf5",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "````{seealso} Vocabulary\n",
    "\n",
    "**Library:** additional coding objects (Ex. functions, data structures) that you can add to your code by importing them\n",
    "\n",
    "````"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3a910ede-8d1d-4e22-8d39-5c23f4414001",
   "metadata": {},
   "source": [
    ":::{admonition} 🌀 More Info: Standard vs. Non-standard Libraries\n",
    ":class: note, dropdown\n",
    "\n",
    "One large chunk of libraries are together referred to as the **Python Standard Library**. These libraries are not a part of the \"core\" language (everything we can do without import statments) but have still been offically accepted into the Python distribution package and are maintained in the same way as the rest of the core language.  You can read more [here](https://docs.python.org/3/tutorial/stdlib.html).\n",
    "\n",
    "There are also libraries which aren't a part of this designation, and are still widely used and developed, but they are simply not included when you hit the \"Download Python\" button.  Non-standard libraries are maintained by groups of developers outside of the Python Software Foundation and anyone can make one.  That includes you!\n",
    "\n",
    ":::"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "00ccb287-2228-4458-9a23-fce6c997264f",
   "metadata": {},
   "source": [
    "## 🐼 Tabular Data and `pandas`"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f25e4306-24da-400e-a53f-b79ae725d19f",
   "metadata": {},
   "source": [
    "**Tabular data** is the type of data that you would see in an excel sheet.  Data is organized into rows (horizontal) and columns (vertical).  It is also common for tabular data to have an index which uniquely lables the rows.  A really common type of tabular data is a CSV file."
   ]
  },
  {
   "attachments": {
    "L3_tabular_data.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "id": "9ce30338-c48b-4190-85d8-ae5a3e74b85d",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "![L3_tabular_data.png](./images/tabular_data.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "adac0d04-ad79-461d-8bc5-6f257a6e9411",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "Where are the rows, columns, and indexes in this excel sheet?"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9b18a11d-b8fe-49eb-884b-981ea489cca5",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "## `pandas` library 🐼"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5ba484cc-3351-4bad-8d54-3c3a552b3507",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "`pandas` is a very common library that is used for working with tabular data.  It is cannonically imported like this:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "83e09185-2537-4598-a169-f98259c0481b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import the libary.  You only have to do this once per file.\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "d9144973-0750-4a70-9150-0c4363e87fc4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# This line shortens the output. It will be explained more later in the notebook.\n",
    "pd.set_option('display.max_rows', 20)\n",
    "pd.set_option('display.max_columns', 12)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5994a5fe-91fa-4c02-846a-aa5f6bf15e16",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "## `pandas` Data structure: `pd.DataFrame`"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6f3b90d4-e2f2-4e19-ad5f-5c96114fdddc",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "A `pandas` **dataframe** is a data structure organized into rows and columns.  Other important parts of the anatomy of a data frame are the columns names and the index.\n",
    "\n",
    "```{image} ./images/pandas_dataframe.png\n",
    ":alt: pandas dataframe\n",
    ":width: 600px\n",
    ":align: center\n",
    "```\n",
    "\n",
    "_Image ammended from \n",
    "[pandas documentation](https://pandas.pydata.org/docs/getting_started/intro_tutorials/01_table_oriented.html)._"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "98aa6b06-4769-429b-b627-7c95088fa144",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "Here's how we can make one from a little bit of scratch data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "4fca4068-885a-4dd7-af32-1a78418b6b3d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Step 1 - Define our fake data with a dictionary\n",
    "forecast_dict = {\n",
    "    'temperature': [25, 20, 27, 31], \n",
    "    'humidity': [60, 65, 62, 80],\n",
    "    'wind_speed': [10, 9, 10, 6]\n",
    "}\n",
    "# Step 2 - use `pd.DataFrame()` to convert our dictionary into a dataframe\n",
    "# Assign it to a variable called `forecast`\n",
    "forecast = pd.DataFrame(forecast_dict, \n",
    "                        index  = ['2020-07-01', '2020-07-02', '2020-07-03', '2020-07-04'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "4823c83a-c224-4088-abbe-b8c1f519ea4a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>temperature</th>\n",
       "      <th>humidity</th>\n",
       "      <th>wind_speed</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2020-07-01</th>\n",
       "      <td>25</td>\n",
       "      <td>60</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-07-02</th>\n",
       "      <td>20</td>\n",
       "      <td>65</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-07-03</th>\n",
       "      <td>27</td>\n",
       "      <td>62</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-07-04</th>\n",
       "      <td>31</td>\n",
       "      <td>80</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            temperature  humidity  wind_speed\n",
       "2020-07-01           25        60          10\n",
       "2020-07-02           20        65           9\n",
       "2020-07-03           27        62          10\n",
       "2020-07-04           31        80           6"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# View the dataframe\n",
    "forecast"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "24483fe8-467e-49b8-be4a-24d2293a5298",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "Our first dataframe!  We can see that we have 3 columns, with column names `temperature`, `humidity`, and `wind_speed`.  We have 4 data entries, or 4 rows, labeled by our indexes, which in this case are just the numbers 0-3."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "34fa3b3e-37a4-4cea-ac30-033d5af8ff7d",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "## Accessing values in `pd.DataFrame()`"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "43802fca-5e6d-4f4a-a0a1-fee6155900ba",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "Even though pandas dataframes are new data structures, some of the patterns for accessing data are very similar to what we saw in the list and dictionary data structures.  What's exciting is that the concept -- indexing -- is conceptually similar and so in response the syntax is also similar.\n",
    "\n",
    "Two main ways you may want to access data in a dataframe might be:\n",
    "1. look at all the data in a single column\n",
    "2. look at all the data in a single row"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e479e307-74c4-4092-8d1d-e57984edee2b",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "### Accessing a column\n",
    "The syntax for this is akin to looking up an item in a dictionary or a value in a list, we use square brackets `[]`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "5b2d6937-14e8-4376-bf2c-375fbcd92d31",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2020-07-01    60\n",
       "2020-07-02    65\n",
       "2020-07-03    62\n",
       "2020-07-04    80\n",
       "Name: humidity, dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forecast['humidity']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e10ec06b-048f-4ec8-89e6-b3f5418b200e",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "### Accessing a row by label\n",
    "\n",
    "The syntax for looking at a particular row is `.loc[]`. It still has square brackets `[]`, as we are seeing over and over again in Python for accessing values, but the new syntax here is the `loc` piece."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "e3a12e14-cc7f-4f33-a9ed-670d4e47b70a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "temperature    20\n",
       "humidity       65\n",
       "wind_speed      9\n",
       "Name: 2020-07-02, dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forecast.loc['2020-07-02']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "966b12b9-47d4-45f8-b186-c616706a2d93",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>temperature</th>\n",
       "      <th>humidity</th>\n",
       "      <th>wind_speed</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2020-07-02</th>\n",
       "      <td>20</td>\n",
       "      <td>65</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-07-03</th>\n",
       "      <td>27</td>\n",
       "      <td>62</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-07-04</th>\n",
       "      <td>31</td>\n",
       "      <td>80</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            temperature  humidity  wind_speed\n",
       "2020-07-02           20        65           9\n",
       "2020-07-03           27        62          10\n",
       "2020-07-04           31        80           6"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forecast.loc['2020-07-02':'2020-07-04']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a8183b1f-dc19-4b0e-9483-72f992b9ff43",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "### Accessing a row by index\n",
    "When looking up data for a row we use the indexes, and the same `[]` syntax that we have seen in lists.  `pandas` uses the `[]` and `:` notation in the exact same way.  The only new addition is the `.iloc` syntax that goes before the square brackets `[]`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "f5e4d6f0-7f26-4ece-9c6d-56e48d5180e9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "temperature    27\n",
       "humidity       62\n",
       "wind_speed     10\n",
       "Name: 2020-07-03, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Display just the row of index 2\n",
    "forecast.iloc[2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "da88c4cf-a1db-4a13-b269-143ff4733a3d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>temperature</th>\n",
       "      <th>humidity</th>\n",
       "      <th>wind_speed</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2020-07-03</th>\n",
       "      <td>27</td>\n",
       "      <td>62</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-07-04</th>\n",
       "      <td>31</td>\n",
       "      <td>80</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            temperature  humidity  wind_speed\n",
       "2020-07-03           27        62          10\n",
       "2020-07-04           31        80           6"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Display rows index 2-3\n",
    "forecast.iloc[2:4]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1142d3f0-94d7-4947-b4fb-1d0080eaf1f3",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "### Summary\n",
    "\n",
    "`.loc[]` and `.iloc[]` may be initially confusing at that is okay. Understanding the difference involves understanding how an index and an index label are different: An index is always a number starting from 0 and sequentially increasing. The label can be any type of text (or number) that references a row. Index labels are often dates/times, which is why pandas is a common tool for timeseries data.\n",
    "\n",
    "While I don't know of any formal or recognized shorthand for `loc` and `iloc`, I sometimes remember them by thinking that `loc` is for \"location\" and `iloc` is for \"integer location\". Calling `iloc` \"integer location\" isn't _strictly_ true 100% of the time, but it's still a helpful mental tool to help me remember which is which.\n",
    "\n",
    "```{image} ./images/accessing_pandas_values.jpg\n",
    ":alt: accessing values in pandas\n",
    ":width: 800px\n",
    ":align: center\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc15c8f1-d61c-4220-8263-eaafb4dbd38f",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "_The checkins below use the `forecast` dataframe we have been working with._"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "d368c3fb-251b-43be-8747-e91230a332d3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>temperature</th>\n",
       "      <th>humidity</th>\n",
       "      <th>wind_speed</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2020-07-01</th>\n",
       "      <td>25</td>\n",
       "      <td>60</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-07-02</th>\n",
       "      <td>20</td>\n",
       "      <td>65</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-07-03</th>\n",
       "      <td>27</td>\n",
       "      <td>62</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2020-07-04</th>\n",
       "      <td>31</td>\n",
       "      <td>80</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            temperature  humidity  wind_speed\n",
       "2020-07-01           25        60          10\n",
       "2020-07-02           20        65           9\n",
       "2020-07-03           27        62          10\n",
       "2020-07-04           31        80           6"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forecast"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "30a381ec-34c4-4fae-9309-f60778afb34b",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    ":::{admonition} Checkin\n",
    ":class: tip\n",
    "\n",
    "1. Write a line of code to get just the `wind_speed` column.\n",
    "2. Write a new line of code to get just the row of index 1.\n",
    "3. Write a line of code to get just the data from July 4th.\n",
    "\n",
    ":::"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "115c98e3-f9ce-4d52-8887-2a253092eab0",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    ":::{admonition} Checkin\n",
    ":class: tip\n",
    "\n",
    "What is the value of `forecast['humidity'].iloc[2]`?\n",
    "\n",
    "a) the row: 27\t62\t10\n",
    "\n",
    "b) the column: 60  65  62  80\n",
    "\n",
    "c) the row: 20\t65\t9\n",
    "\n",
    "d) 65\n",
    "\n",
    "e) 62\n",
    "\n",
    "[Poll link](https://PollEv.com/multiple_choice_polls/zKsN4K3CsQzZu1fLI0qQW/respond)\n",
    "\n",
    ":::"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "49f3773d-7d39-45d9-b41b-50bd0fb56e40",
   "metadata": {},
   "source": [
    "### 🌟 A note about assigning dataframes to variables\n",
    "It's easy to feel like when you do something to find one part of a dataframe (Ex. `forecast.iloc[2:4]` or `forecast['humidity']`) you have changed the dataframe.  This is not the case, however, and the original dataframe stays the same."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8e1551d9-fa78-4e3b-96d8-929b28c10c3a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Looking up the 2nd and 3rd indices of the forecast dataframe\n",
    "forecast.iloc[2:4]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "26db2584-9ef0-44d2-95a1-a6cb31ab54d2",
   "metadata": {},
   "outputs": [],
   "source": [
    "# `forecast` is still the same dataframe\n",
    "forecast"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4f460c35-21e3-4ad4-a4ab-7dc94f5a699d",
   "metadata": {},
   "source": [
    "This is nice for data exploration, when you don't want the values to change every time you look at them.  If you are trying to permanently change your data, you need to remember to update the dataframe by reassigning the variable."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "687f319d-c00f-4c87-b7d1-ec1da1a4515b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Updating the forecast variable with just the two rows\n",
    "forecast = forecast.iloc[2:4]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c2fa2e97-5403-4a21-97e6-a7d8348845a6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# `forecast was updated`\n",
    "forecast"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1bb877ad-932c-4447-bdb3-dc64fb3071fb",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "## Using some real data\n",
    "\n",
    "Part of accessing data is informing your code where it can find that data. In this lesson we are going to access data that is located in the efs shared folder of the SARP drive."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "e4484f0c-6e27-4349-9e51-b92e1dbef639",
   "metadata": {},
   "outputs": [],
   "source": [
    "from pathlib import Path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "fcc49e8a-e238-49a3-8ab2-d7f6a4c0000e",
   "metadata": {},
   "outputs": [],
   "source": [
    "DATADIR = Path('/efs/sarp/data/rawdata_readonly')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "2b813f38-c034-4500-b78e-6556dd26b7d9",
   "metadata": {},
   "outputs": [],
   "source": [
    "file = (\n",
    "    DATADIR\n",
    "    / 'flight-test-data'\n",
    "    / 'discoveraq-mrg60-p3b_merge_20110701_R4.ict'\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f2e4b84c-9641-4be4-b16c-7ff1a847e287",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "`pd.read_csv()` is one of the go-to data opening functions in pandas. The first argument is the filepath and in a very clean file that may be the only argument you give. Often, though, you have to give additional arguments to help pandas read your file. `skiprows` is a common one. It tells pandas to ignore that number of rows at the top of the file. Rows at the top often contain metadata information."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "b6c04992-992e-4100-a923-542fa1120622",
   "metadata": {},
   "outputs": [],
   "source": [
    "discover = pd.read_csv(file, skiprows=171)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "0fd305a3-f58c-45ba-9868-ec1ef2542737",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([' UTC', ' JDAY', ' INDEX', ' FLIGHT', ' LOCAL_SUN_TIME',\n",
       "       ' LATITUDE', ' LONGITUDE', ' ALTP', ' PRESSURE', ' TEMPERATURE',\n",
       "       ' THETA', ' O3COLUMN', ' SZA', ' WNS', ' WND', ' FMS_TAS',\n",
       "       ' FMS_SAT', ' FMS_GRD_SPD', ' Heading', ' FMS_TRK', ' IRS_PITCH',\n",
       "       ' IRS_ROLL', ' IRS_VERT_ACC', ' ADC_IAS', ' GPS_ALT',\n",
       "       ' A_DewPoint', ' A_CabinPressure', ' A_SurfTemp', ' A_TotalTemp',\n",
       "       ' A_JNO2_Nadir', ' A_JNO2_Zenith', ' A_RadarAlt',\n",
       "       ' C_DiffPressure', ' C_PotTempDegK', ' C_MachNumber',\n",
       "       ' C_CabAltitude', ' C_VaporPresWater', ' C_SatVaporPresWater',\n",
       "       ' C_SatVaporPresIce', ' C_MixingRatio', ' C_RelHumidity',\n",
       "       ' ProfileNumber', ' SiteSeqNumber', ' MissedApproachNumber',\n",
       "       ' ProfileCenterBearing', ' H2O(DLH)', ' NO', ' NOy', ' NO2_NCAR',\n",
       "       ' O3', ' Carbon_Monoxide_mixing_ratio', ' Methane_mixing_ratio',\n",
       "       ' Carbon_dioxide_mixing_ratio', ' NO2_LIF', ' PNs_TD-LIF',\n",
       "       ' ANs_TD-LIF', ' HNO3_TD-LIF', ' CH2O_DFGAS', ' Abs470tot',\n",
       "       ' Abs532tot', ' Abs660tot', ' nAPS_total', ' sAPS_total',\n",
       "       ' vAPS_total', ' nAPS_fine(Daero_lt_1.0um)',\n",
       "       ' sAPS_fine(Daero_lt_1.0um)', ' vAPS_fine(Daero_lt_1.0um)',\n",
       "       ' nAPS_coarse(Daero_ge_1.0um)', ' sAPS_coarse(Daero_ge_1.0um)',\n",
       "       ' vAPS_coarse(Daero_ge_1.0um)', ' BlackCarbonMassConcentration',\n",
       "       ' CN>3nm', ' CN>10nm', ' nonvolCN>10nm', ' Scat450tot',\n",
       "       ' Scat550tot', ' Scat700tot', ' Scat450sub', ' Scat550sub',\n",
       "       ' Scat700sub', ' nUHSAS(60-1000_nm)', ' sUHSAS(60-1000_nm)',\n",
       "       ' vUHSAS(60-1000_nm)', ' nLAS(0.9-7.7_um)', ' sLAS(0.9-7.7_um)',\n",
       "       ' vLAS(0.9-7.7_um)',\n",
       "       ' Angstrom_Exponent_of_Scattering_at_450and700nm',\n",
       "       ' Angstrom_Exponent_of_Scattering_at_450and550nm',\n",
       "       ' Angstrom_Exponent_of_Absorption_at_450and700nm',\n",
       "       ' Angstrom_Exponent_of_Absorption_at_450and550nm',\n",
       "       ' SingleScatteringAlbedo_at_450nm',\n",
       "       ' SingleScatteringAlbedo_at_550nm',\n",
       "       ' SingleScatteringAlbedo_at_700nm',\n",
       "       ' SingleScatteringAlbedo_at_550nm(ambient)', ' nSMPS', ' sSMPS',\n",
       "       ' vSMPS', ' Acetaldehyde', ' Acetonitrile', ' Acetone',\n",
       "       ' Methanol', ' Toluene', ' Isoprene', ' MVK_MAC', ' Xylenes_ETB',\n",
       "       ' Monoterpenes', ' RHamb', ' RHdry', ' RHwet', ' SCdry', ' SCwet',\n",
       "       ' gamma', ' fRH_80_20', ' SCamb', ' EXTamb532', ' EXTdry532',\n",
       "       ' SCamb532', ' SCdry532', ' ABSdry532', ' Chloride', ' Nitrite',\n",
       "       ' Nitrate', ' Sulfate', ' Sodium', ' Ammonium', ' Potassium',\n",
       "       ' Magnesium', ' Calcium', ' Water-Soluble_Organic_Carbon_Mass'],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discover.columns.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "5a6655f8-3c89-45b9-bf01-29e20bd2e759",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>UTC</th>\n",
       "      <th>JDAY</th>\n",
       "      <th>INDEX</th>\n",
       "      <th>FLIGHT</th>\n",
       "      <th>LOCAL_SUN_TIME</th>\n",
       "      <th>LATITUDE</th>\n",
       "      <th>...</th>\n",
       "      <th>Sodium</th>\n",
       "      <th>Ammonium</th>\n",
       "      <th>Potassium</th>\n",
       "      <th>Magnesium</th>\n",
       "      <th>Calcium</th>\n",
       "      <th>Water-Soluble_Organic_Carbon_Mass</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>51930</td>\n",
       "      <td>182</td>\n",
       "      <td>10001</td>\n",
       "      <td>1</td>\n",
       "      <td>9.403683</td>\n",
       "      <td>37.989785</td>\n",
       "      <td>...</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>51990</td>\n",
       "      <td>182</td>\n",
       "      <td>10002</td>\n",
       "      <td>1</td>\n",
       "      <td>9.421592</td>\n",
       "      <td>37.988955</td>\n",
       "      <td>...</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>52050</td>\n",
       "      <td>182</td>\n",
       "      <td>10003</td>\n",
       "      <td>1</td>\n",
       "      <td>9.436436</td>\n",
       "      <td>37.949315</td>\n",
       "      <td>...</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>52110</td>\n",
       "      <td>182</td>\n",
       "      <td>10004</td>\n",
       "      <td>1</td>\n",
       "      <td>9.450309</td>\n",
       "      <td>37.923548</td>\n",
       "      <td>...</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>52170</td>\n",
       "      <td>182</td>\n",
       "      <td>10005</td>\n",
       "      <td>1</td>\n",
       "      <td>9.466201</td>\n",
       "      <td>37.962125</td>\n",
       "      <td>...</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>423</th>\n",
       "      <td>77310</td>\n",
       "      <td>182</td>\n",
       "      <td>10424</td>\n",
       "      <td>1</td>\n",
       "      <td>16.450163</td>\n",
       "      <td>38.157933</td>\n",
       "      <td>...</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>1.141667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>424</th>\n",
       "      <td>77370</td>\n",
       "      <td>182</td>\n",
       "      <td>10425</td>\n",
       "      <td>1</td>\n",
       "      <td>16.468866</td>\n",
       "      <td>38.108382</td>\n",
       "      <td>...</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>1.145000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>425</th>\n",
       "      <td>77430</td>\n",
       "      <td>182</td>\n",
       "      <td>10426</td>\n",
       "      <td>1</td>\n",
       "      <td>16.487703</td>\n",
       "      <td>38.061325</td>\n",
       "      <td>...</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>426</th>\n",
       "      <td>77490</td>\n",
       "      <td>182</td>\n",
       "      <td>10427</td>\n",
       "      <td>1</td>\n",
       "      <td>16.505194</td>\n",
       "      <td>38.013657</td>\n",
       "      <td>...</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>427</th>\n",
       "      <td>77550</td>\n",
       "      <td>182</td>\n",
       "      <td>10428</td>\n",
       "      <td>1</td>\n",
       "      <td>16.520896</td>\n",
       "      <td>37.972985</td>\n",
       "      <td>...</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.0</td>\n",
       "      <td>-999999.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>428 rows × 129 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       UTC   JDAY   INDEX   FLIGHT   LOCAL_SUN_TIME   LATITUDE  ...   Sodium  \\\n",
       "0    51930    182   10001        1         9.403683  37.989785  ...  -999999   \n",
       "1    51990    182   10002        1         9.421592  37.988955  ...  -999999   \n",
       "2    52050    182   10003        1         9.436436  37.949315  ...  -999999   \n",
       "3    52110    182   10004        1         9.450309  37.923548  ...  -999999   \n",
       "4    52170    182   10005        1         9.466201  37.962125  ...  -999999   \n",
       "..     ...    ...     ...      ...              ...        ...  ...      ...   \n",
       "423  77310    182   10424        1        16.450163  38.157933  ...  -999999   \n",
       "424  77370    182   10425        1        16.468866  38.108382  ...  -999999   \n",
       "425  77430    182   10426        1        16.487703  38.061325  ...  -999999   \n",
       "426  77490    182   10427        1        16.505194  38.013657  ...  -999999   \n",
       "427  77550    182   10428        1        16.520896  37.972985  ...  -999999   \n",
       "\n",
       "      Ammonium   Potassium   Magnesium   Calcium  \\\n",
       "0      -999999   -999999.0   -999999.0 -999999.0   \n",
       "1      -999999   -999999.0   -999999.0 -999999.0   \n",
       "2      -999999   -999999.0   -999999.0 -999999.0   \n",
       "3      -999999   -999999.0   -999999.0 -999999.0   \n",
       "4      -999999   -999999.0   -999999.0 -999999.0   \n",
       "..         ...         ...         ...       ...   \n",
       "423    -999999   -999999.0   -999999.0 -999999.0   \n",
       "424    -999999   -999999.0   -999999.0 -999999.0   \n",
       "425    -999999   -999999.0   -999999.0 -999999.0   \n",
       "426    -999999   -999999.0   -999999.0 -999999.0   \n",
       "427    -999999   -999999.0   -999999.0 -999999.0   \n",
       "\n",
       "      Water-Soluble_Organic_Carbon_Mass  \n",
       "0                        -999999.000000  \n",
       "1                        -999999.000000  \n",
       "2                        -999999.000000  \n",
       "3                        -999999.000000  \n",
       "4                        -999999.000000  \n",
       "..                                  ...  \n",
       "423                            1.141667  \n",
       "424                            1.145000  \n",
       "425                      -999999.000000  \n",
       "426                      -999999.000000  \n",
       "427                      -999999.000000  \n",
       "\n",
       "[428 rows x 129 columns]"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discover"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "44da43f7-39f3-4b8c-9804-c1cab630f53f",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    ":::{attention} \n",
    "Practice Problem #2\n",
    ":::"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7ed4c7b-86b8-48de-b0a1-1bd436419a73",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "## Functions/methods for pandas dataframes\n",
    "\n",
    "The dataframe represents the data but functions or methods allow us to manipulate that data. Let's look at a few examples using a single column of the `discover` dataframe, the relative humidity."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "a4a8ccad-ce5e-4df9-8fc5-f1a2f7a6daea",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      50.204500\n",
       "1      54.015667\n",
       "2      47.659000\n",
       "3      47.801167\n",
       "4      52.815000\n",
       "         ...    \n",
       "423    71.283167\n",
       "424    65.042333\n",
       "425    64.859500\n",
       "426    67.741333\n",
       "427    70.299500\n",
       "Name:  RHamb, Length: 428, dtype: float64"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discover[' RHamb']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "58221ee3-5788-4a0d-be4c-7b572bf414eb",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "Some aggregations:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "fdf2d4c6-223b-4b5c-b842-a30d48290208",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "36.807866043544394"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discover[' RHamb'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "60ed12fe-cfcd-4040-95c4-d76a29b7d05f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "18.736073018487495"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discover[' RHamb'].std()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c77e68d8-b90b-4f13-ad18-659ddb774829",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    ":::{admonition} Notice\n",
    ":class: note\n",
    "The syntax for methods on pandas dataframes is the same one that we saw for lists. It is `df.method()`, just like `list.method()`.\n",
    ":::"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "96875266-5bf4-42a1-bea3-084666d13ff0",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "We can use comparisons to check values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "82c2ca88-80a0-49bc-8557-8abe224f92ff",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       True\n",
       "1       True\n",
       "2      False\n",
       "3      False\n",
       "4       True\n",
       "       ...  \n",
       "423     True\n",
       "424     True\n",
       "425     True\n",
       "426     True\n",
       "427     True\n",
       "Name:  RHamb, Length: 428, dtype: bool"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discover[' RHamb'] > 50"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "611aa704-488d-42ba-a9f0-5038c5a0838b",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "Using a comparison together with a `.sum()` gives us the number of rows that meet the condition."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "id": "87f007d9-e50b-4de5-a4e7-5eed07c7687c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "108"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rd_over_50 = discover[' RHamb'] > 50\n",
    "rd_over_50.sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b4b799bc-9bf2-44cc-8c13-341659df390f",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "Let's take a look at another column, CO."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "0a23ec40-99fc-4534-9a78-f45f2850f2a4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      144.336400\n",
       "1      146.655167\n",
       "2      143.108367\n",
       "3      144.393611\n",
       "4      143.252833\n",
       "          ...    \n",
       "423    143.836500\n",
       "424    144.404167\n",
       "425    143.722571\n",
       "426    146.834500\n",
       "427    152.337167\n",
       "Name:  Carbon_Monoxide_mixing_ratio, Length: 428, dtype: float64"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discover[' Carbon_Monoxide_mixing_ratio']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "id": "942e09e7-a88c-4009-94d9-0e13a34f0074",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-11550.41168590564"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discover[' Carbon_Monoxide_mixing_ratio'].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "15163f95-497d-4033-87d3-cfa5101fbb62",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "A negative value? Why might that be? If we look back at the data we see that there are very large negative numbers (-999999) that likely represent nodata values.\n",
    "\n",
    "One way to confirm this is the case would be to sort the rows. The method for this is `.sort_values()`. It's worth taking a moment to glance at [the documentation page](https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.sort_values.html) for `.sort_values()`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "id": "9e6b779c-6fe5-4f37-9868-59e30891c65d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "310   -999999.000000\n",
       "187   -999999.000000\n",
       "311   -999999.000000\n",
       "188   -999999.000000\n",
       "26    -999999.000000\n",
       "           ...      \n",
       "249       167.402368\n",
       "267       169.639000\n",
       "254       169.929500\n",
       "387       170.772500\n",
       "386       170.895000\n",
       "Name:  Carbon_Monoxide_mixing_ratio, Length: 428, dtype: float64"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discover[' Carbon_Monoxide_mixing_ratio'].sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "b0337e8e-2453-4ca1-84bb-65f3ea0dda4b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "386       170.895000\n",
       "387       170.772500\n",
       "254       169.929500\n",
       "267       169.639000\n",
       "249       167.402368\n",
       "           ...      \n",
       "26    -999999.000000\n",
       "187   -999999.000000\n",
       "188   -999999.000000\n",
       "310   -999999.000000\n",
       "311   -999999.000000\n",
       "Name:  Carbon_Monoxide_mixing_ratio, Length: 428, dtype: float64"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discover[' Carbon_Monoxide_mixing_ratio'].sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3ca4b369-a092-4a4f-81b9-380c01466b52",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "### A long chain of commands\n",
    "\n",
    "With pandas you can string several commands together. This is a nice feature, but can make reading other people's code intimidating sometimes. If you see a long string step back and tackle reading it one method at a time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "e43b195b-3d17-4d09-9b6f-314d55bb91a8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.2505854800936768"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(discover[' RHamb'] > 50).sort_values(ascending=False).drop(0).mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "87a5f361-6eba-4949-8e94-ac51b8204479",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    ":::{attention} \n",
    "Practice Problem #3\n",
    ":::"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eab3c9f7-629f-491b-8c2e-73e853cc3f4a",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    "## Closing Poll\n",
    "\n",
    "[The closing poll link](https://PollEv.com/clickable_images/WHersUlQe5SsG68KOHyfL/respond) which is, as all the others are, anonymous."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "03d8a4cb-5c19-4192-b267-48e68839fd6f",
   "metadata": {
    "user_expressions": []
   },
   "source": [
    ":::{danger} Shutdown \n",
    "Please shut down your server! (File > Hub Control Panel) \n",
    ":::"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f474406e-5f19-4b51-a5f4-2ce135aa0619",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:notebook] *",
   "language": "python",
   "name": "conda-env-notebook-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}