nbs/04_intro_pandas.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](../images/logos.jpg \"MiCMOR, KIT Campus Alpin\")\n",
    "\n",
    "**[MiCMOR](https://micmor.kit.edu) [SummerSchool \"Environmental Data Science: From Data Exploration to Deep Learning\"](https://micmor.kit.edu/sites/default/files/MICMoR%20Summer%20School%202019%20Flyer.pdf)**  \n",
    "IMK-IFU KIT Campus Alpin, Sept. 4 - 13 2019, Garmisch-Partenkirchen, Germany.\n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Pandas\n",
    "\n",
    "![](https://img.etimg.com/thumb/msid-51641992,width-643,imgsize-288661,resizemode-4/kung-fu-panda-3-review-film-ticks-all-the-boxes.jpg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Working with tabular data\n",
    "\n",
    "The Pandas package is for many people *the reason* for much of the recent success of python as the Data Science language.\n",
    "\n",
    "In a  Stackoverflow [study](https://stackoverflow.blog/2017/09/14/python-growing-quickly/) it was concluded that:\n",
    ">These analyses suggest two conclusions. First, the fastest-growing use of Python is for data science, machine learning and academic research. This is particularly visible in the growth of the pandas package, which is the fastest-growing Python-related tag on the site. As for which industries are using Python, we found that it is more visited in a few industries, such as electronics, manufacturing, software, government, and especially universities. However, Python’s growth is spread pretty evenly across industries. In combination this tells a story of data science and machine learning becoming more common in many types of companies, and Python becoming a common choice for that purpose.\n",
    "\n",
    "<center>\n",
    "<img width=\"400\" src=https://1015711.v1.pressablecdn.com/wp-content/uploads/2017/09/related_tags_over_time-1-1200x1200.png>\n",
    "<p><center><em>The rise of pandas</em></center></p>\n",
    "</center>\n",
    "\n",
    "So we first explore the basics of working with pandas a bit and later use it for some analysis.\n",
    "\n",
    "### Basic data types: Seris, DataFrame\n",
    "\n",
    "In general we import pandas like this. Usually you also want numpy for some processing so we import it too:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pylab as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pandas features two basic types:  **Series** and **DataFrame** ([see also the relevant section in the getting started guide](https://pandas.pydata.org/pandas-docs/stable/getting_started/dsintro.html)). \n",
    "A series holds an one-dimensional labeled array, a dataframe is a 2-dimensional labeled data structure with columns of potentially different types (think spreadsheet).\n",
    "\n",
    "Both are composed of data and an index (the row labels). A dataframe has columns wioth column labels.\n",
    "\n",
    "Here is how they are defined and what they look like.\n",
    "\n",
    "First a **Series**:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    0\n",
       "1    9\n",
       "2    8\n",
       "3    0\n",
       "4    5\n",
       "5    9\n",
       "6    5\n",
       "7    3\n",
       "8    7\n",
       "9    2\n",
       "Name: Flux1, dtype: int64"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s = pd.Series(np.random.randint(0, 10, size=10), name='Flux1')\n",
    "s"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If the index is ommited, the Series gets an integer index (you can set it as an argument). Usually, you also want to add a name to describe your data.\n",
    "\n",
    "Access the values with `.values`, the index with `.index`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Values [0 9 8 0 5 9 5 3 7 2]\n",
      "Index RangeIndex(start=0, stop=10, step=1)\n"
     ]
    }
   ],
   "source": [
    "print('Values', s.values)\n",
    "print('Index', s.index)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Another example where you set an index expicitly (but no name):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "a   -0.456086\n",
       "b    0.010117\n",
       "c    0.360907\n",
       "d    1.058469\n",
       "e   -0.678062\n",
       "dtype: float64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s = pd.Series(np.random.randn(5), index=['a', 'b', 'c', 'd', 'e'])\n",
    "s"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note the dtype. This is the data type. This can be `object` for any non numeric data (i.e. strings) or int, float for numeric data. There's also a special dtype (categorical) that we will have a look at later. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we look at a **DataFrames**:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You'll probably more often work with `DataFrames`, as they can store multiple variables.\n",
    "\n",
    "You can create dataframes in many ways, but most often you will read the data from file (i.e. with the read_csv() function).\n",
    "\n",
    "One way to define a dataframe is with a dictionary of columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>one</th>\n",
       "      <th>two</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>4.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>3.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>2.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1.3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   one  two\n",
       "0    1  4.1\n",
       "1    2  3.4\n",
       "2    3  2.1\n",
       "3    4  1.3"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d = {'one': [1,2,3,4], 'two': [4.1, 3.4, 2.1, 1.3]}\n",
    "\n",
    "pd.DataFrame(d)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Creating the dataframe while passing a custom index:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>one</th>\n",
       "      <th>two</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>1</td>\n",
       "      <td>4.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>2</td>\n",
       "      <td>3.4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>3</td>\n",
       "      <td>2.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d</th>\n",
       "      <td>4</td>\n",
       "      <td>1.3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   one  two\n",
       "a    1  4.1\n",
       "b    2  3.4\n",
       "c    3  2.1\n",
       "d    4  1.3"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(d, index=['a', 'b', 'c', 'd'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Or by passing Series elements:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>one</th>\n",
       "      <th>two</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d</th>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   one  two\n",
       "a  1.0  1.0\n",
       "b  2.0  2.0\n",
       "c  3.0  3.0\n",
       "d  NaN  4.0"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d = {'one': pd.Series([1., 2., 3.], index=['a', 'b', 'c']),\n",
    "     'two': pd.Series([1., 2., 3., 4.], index=['a', 'b', 'c', 'd'])}\n",
    "\n",
    "pd.DataFrame(d)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's look at the results. Pandas matches data by the index. Thus, column one has now a missing value `NaN` in row d.\n",
    "\n",
    "This also happens if you specify columns that do not exist in the given dictionary:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>two</th>\n",
       "      <th>three</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>d</th>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   two three\n",
       "d  4.0   NaN\n",
       "b  2.0   NaN\n",
       "a  1.0   NaN"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(d, index=['d', 'b', 'a'], columns=['two', 'three'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can access (and set) the column and row labels like this:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['a', 'b', 'c', 'd'], dtype='object')"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.DataFrame(d)\n",
    "df.index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['one', 'two'], dtype='object')"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>V1</th>\n",
       "      <th>V2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d</th>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    V1   V2\n",
       "a  1.0  1.0\n",
       "b  2.0  2.0\n",
       "c  3.0  3.0\n",
       "d  NaN  4.0"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# setting new column names\n",
    "df2 = df.copy()\n",
    "df2.columns = ['V1', 'V2']\n",
    "df2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can create new columns like this:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "df3 = df.copy()\n",
    "df3['three'] = df3['one'] * df3['two']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note, there is also a differnt syntax to access columns that is very common. \n",
    "\n",
    "To access column 'two' like this:\n",
    "```python\n",
    "df.two\n",
    "# is the same as\n",
    "df['two']\n",
    "```\n",
    "\n",
    "However, the first version only works if the column names does not contain spaces, '.' or other symbols that might confuse the python language parser. Thus, do not use points in your column names.\n",
    "\n",
    "You can also only use this syntax on the right side of an assignment (see [here](https://pandas.pydata.org/pandas-docs/stable/indexing.html#attribute-access))."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# this works\n",
    "df4 = df.copy()\n",
    "df4['three'] = df4.one * df4.two"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.6/site-packages/ipykernel_launcher.py:3: UserWarning: Pandas doesn't allow columns to be created via a new attribute name - see https://pandas.pydata.org/pandas-docs/stable/indexing.html#attribute-access\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n"
     ]
    },
    {
     "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>one</th>\n",
       "      <th>two</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d</th>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   one  two\n",
       "a  1.0  1.0\n",
       "b  2.0  2.0\n",
       "c  3.0  3.0\n",
       "d  NaN  4.0"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# this does not work\n",
    "df5 = df.copy()\n",
    "df5.three = df5.one * df5.two\n",
    "df5"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To delete a column you can either use `del` or `drop(colname, axis=1)` (axis=0 in pandas refers to rows-wise operations, axis=1 to column-wise operations):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>two</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d</th>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   two\n",
       "a  1.0\n",
       "b  2.0\n",
       "c  3.0\n",
       "d  4.0"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "del df5['one']\n",
    "df5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>one</th>\n",
       "      <th>three</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>2.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>3.0</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   one  three\n",
       "a  1.0    1.0\n",
       "b  2.0    4.0\n",
       "c  3.0    9.0\n",
       "d  NaN    NaN"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>three</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d</th>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   three\n",
       "a    1.0\n",
       "b    4.0\n",
       "c    9.0\n",
       "d    NaN"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# drop one row\n",
    "display(df4.drop('two', axis=1))\n",
    "\n",
    "# drop multiple rows\n",
    "df4.drop(['one','two'], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Indexing and selection\n",
    "\n",
    "The basics of indexing are as follows:\n",
    "\n",
    "| Operation                      |Syntax        |Result    |\n",
    "|--------------------------------|--------------|----------|\n",
    "| Select column                  |df[col]       |Series    |\n",
    "| Select row by label            |df.loc[label] |Series    |\n",
    "| Select row by integer location |df.iloc[loc]  |Series    |\n",
    "| Slice rows                     |df[5:10]      |DataFrame |\n",
    "| Select rows by boolean vector  |df[bool_vec]  |DataFrame |\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "one    1.0\n",
       "two    1.0\n",
       "Name: a, dtype: float64"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Examples\n",
    "\n",
    "# 1) get first row (use iloc)\n",
    "df.iloc[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "a    1.0\n",
       "b    2.0\n",
       "c    3.0\n",
       "d    NaN\n",
       "Name: one, dtype: float64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 2) get all rows and first col (use iloc)\n",
    "df.iloc[:,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "one    2.0\n",
       "two    2.0\n",
       "Name: b, dtype: float64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 3) select row by label\n",
    "df.loc['b']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "a    1.0\n",
       "b    2.0\n",
       "c    3.0\n",
       "d    4.0\n",
       "Name: two, dtype: float64"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 4) select column by label\n",
    "df['two']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>one</th>\n",
       "      <th>two</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d</th>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   one  two\n",
       "c  3.0  3.0\n",
       "d  NaN  4.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>one</th>\n",
       "      <th>two</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d</th>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   one  two\n",
       "c  3.0  3.0\n",
       "d  NaN  4.0"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 5) select rows 2 to 4\n",
    "display(df.iloc[2:4])\n",
    "df[2:4] # the same"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## More advanced data selection and filtering\n",
    "\n",
    "A nice and very useful feature is the following syntax. Selecting all rows where a certain value in a column fulfills a certain criterium."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>sensor1</th>\n",
       "      <th>sensor2</th>\n",
       "      <th>experiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.649669</td>\n",
       "      <td>0.682228</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.592045</td>\n",
       "      <td>0.888380</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.402215</td>\n",
       "      <td>0.326458</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.302534</td>\n",
       "      <td>0.166441</td>\n",
       "      <td>b</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.130584</td>\n",
       "      <td>0.811590</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.068118</td>\n",
       "      <td>0.594521</td>\n",
       "      <td>b</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.608773</td>\n",
       "      <td>0.965776</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>-0.163694</td>\n",
       "      <td>0.933927</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-0.498306</td>\n",
       "      <td>0.356632</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.193836</td>\n",
       "      <td>0.157421</td>\n",
       "      <td>b</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    sensor1   sensor2 experiment\n",
       "0  0.649669  0.682228          a\n",
       "1  1.592045  0.888380          a\n",
       "2  1.402215  0.326458          a\n",
       "3 -0.302534  0.166441          b\n",
       "4 -1.130584  0.811590          a\n",
       "5  1.068118  0.594521          b\n",
       "6  1.608773  0.965776          c\n",
       "7 -0.163694  0.933927          c\n",
       "8 -0.498306  0.356632          a\n",
       "9  0.193836  0.157421          b"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# some fake data\n",
    "data = {'sensor1': np.random.randn(10), \n",
    "        'sensor2': np.random.random(10),\n",
    "        'experiment': ['a','a','a','b','a','b','c','c','a','b']}\n",
    "df = pd.DataFrame(data)\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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>sensor1</th>\n",
       "      <th>sensor2</th>\n",
       "      <th>experiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.649669</td>\n",
       "      <td>0.682228</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.592045</td>\n",
       "      <td>0.888380</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.130584</td>\n",
       "      <td>0.811590</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.068118</td>\n",
       "      <td>0.594521</td>\n",
       "      <td>b</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.608773</td>\n",
       "      <td>0.965776</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>-0.163694</td>\n",
       "      <td>0.933927</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    sensor1   sensor2 experiment\n",
       "0  0.649669  0.682228          a\n",
       "1  1.592045  0.888380          a\n",
       "4 -1.130584  0.811590          a\n",
       "5  1.068118  0.594521          b\n",
       "6  1.608773  0.965776          c\n",
       "7 -0.163694  0.933927          c"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get all rows where sensor2 recorded values > 0.8\n",
    "df[df.sensor2 > 0.5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>sensor1</th>\n",
       "      <th>sensor2</th>\n",
       "      <th>experiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.649669</td>\n",
       "      <td>0.682228</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.592045</td>\n",
       "      <td>0.888380</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.402215</td>\n",
       "      <td>0.326458</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.130584</td>\n",
       "      <td>0.811590</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-0.498306</td>\n",
       "      <td>0.356632</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    sensor1   sensor2 experiment\n",
       "0  0.649669  0.682228          a\n",
       "1  1.592045  0.888380          a\n",
       "2  1.402215  0.326458          a\n",
       "4 -1.130584  0.811590          a\n",
       "8 -0.498306  0.356632          a"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get all rows of group a\n",
    "df[df.experiment == 'a']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "More complex queries can be \"chained\", combined into one binary selection or achieved using `query`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "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>sensor1</th>\n",
       "      <th>sensor2</th>\n",
       "      <th>experiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-0.498306</td>\n",
       "      <td>0.356632</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    sensor1   sensor2 experiment\n",
       "8 -0.498306  0.356632          a"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Task: get all rows of group a and with sensor 1 between -0.5 and 0.5\n",
    "    \n",
    "# chained\n",
    "df.loc[df.experiment == 'a'].loc[df.sensor1 > -0.5].loc[df.sensor1 < 0.5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>sensor1</th>\n",
       "      <th>sensor2</th>\n",
       "      <th>experiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-0.498306</td>\n",
       "      <td>0.356632</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    sensor1   sensor2 experiment\n",
       "8 -0.498306  0.356632          a"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# combined (better)\n",
    "df[(df.experiment == 'a') & (df.sensor1 > -0.5) & (df.sensor1 < 0.5)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "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>sensor1</th>\n",
       "      <th>sensor2</th>\n",
       "      <th>experiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-0.498306</td>\n",
       "      <td>0.356632</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    sensor1   sensor2 experiment\n",
       "8 -0.498306  0.356632          a"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# using query (elegant, but actually not as common since it's a bit unpythonic (more like SQL)...\n",
    "df.query('experiment == \"a\" and sensor1 > -0.5 and sensor1 < 0.5')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can also test for elements in a list with `.isin`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "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>sensor1</th>\n",
       "      <th>sensor2</th>\n",
       "      <th>experiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.649669</td>\n",
       "      <td>0.682228</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.592045</td>\n",
       "      <td>0.888380</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.402215</td>\n",
       "      <td>0.326458</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.130584</td>\n",
       "      <td>0.811590</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.608773</td>\n",
       "      <td>0.965776</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>-0.163694</td>\n",
       "      <td>0.933927</td>\n",
       "      <td>c</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-0.498306</td>\n",
       "      <td>0.356632</td>\n",
       "      <td>a</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    sensor1   sensor2 experiment\n",
       "0  0.649669  0.682228          a\n",
       "1  1.592045  0.888380          a\n",
       "2  1.402215  0.326458          a\n",
       "4 -1.130584  0.811590          a\n",
       "6  1.608773  0.965776          c\n",
       "7 -0.163694  0.933927          c\n",
       "8 -0.498306  0.356632          a"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.experiment.isin(['a','c'])]\n",
    "# this is actually the same as df[df.experiment != 'b'] in this case"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Groupby operations\n",
    "\n",
    "This is another super power of pandas: grouped operations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "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>sensor1</th>\n",
       "      <th>sensor2</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>experiment</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>0.403008</td>\n",
       "      <td>0.613058</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>0.319807</td>\n",
       "      <td>0.306128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>0.722540</td>\n",
       "      <td>0.949851</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             sensor1   sensor2\n",
       "experiment                    \n",
       "a           0.403008  0.613058\n",
       "b           0.319807  0.306128\n",
       "c           0.722540  0.949851"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# average results by experiment\n",
    "df.groupby('experiment').mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can iterate through groups too..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('a',     sensor1   sensor2 experiment\n",
      "0  0.649669  0.682228          a\n",
      "1  1.592045  0.888380          a\n",
      "2  1.402215  0.326458          a\n",
      "4 -1.130584  0.811590          a\n",
      "8 -0.498306  0.356632          a)\n",
      "('b',     sensor1   sensor2 experiment\n",
      "3 -0.302534  0.166441          b\n",
      "5  1.068118  0.594521          b\n",
      "9  0.193836  0.157421          b)\n",
      "('c',     sensor1   sensor2 experiment\n",
      "6  1.608773  0.965776          c\n",
      "7 -0.163694  0.933927          c)\n"
     ]
    }
   ],
   "source": [
    "grouped = df.groupby('experiment')\n",
    "for g in grouped:\n",
    "    print(g)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are many more possibilities, but this should give you a glimpse into the powers of pandas! For sure one reason why it became so popular in Data Science..."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Combining and merging datasets\n",
    "### Database-Style DataFrame Joins"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "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>key</th>\n",
       "      <th>data1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>b</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>b</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>a</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>c</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>a</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>a</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>b</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  key  data1\n",
       "0   b      0\n",
       "1   b      1\n",
       "2   a      2\n",
       "3   c      3\n",
       "4   a      4\n",
       "5   a      5\n",
       "6   b      6"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>key</th>\n",
       "      <th>data2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>a</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>b</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>d</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  key  data2\n",
       "0   a      0\n",
       "1   b      1\n",
       "2   d      2"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df1 = pd.DataFrame({'key': ['b', 'b', 'a', 'c', 'a', 'a', 'b'],\n",
    "                    'data1': range(7)})\n",
    "df2 = pd.DataFrame({'key': ['a', 'b', 'd'],\n",
    "                    'data2': range(3)})\n",
    "display(df1)\n",
    "display(df2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now merge dataframe `df1` and `df2`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "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>key</th>\n",
       "      <th>data1</th>\n",
       "      <th>data2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>b</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>b</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>b</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>a</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>a</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>a</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  key  data1  data2\n",
       "0   b      0      1\n",
       "1   b      1      1\n",
       "2   b      6      1\n",
       "3   a      2      0\n",
       "4   a      4      0\n",
       "5   a      5      0"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.merge(df1, df2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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>key</th>\n",
       "      <th>data1</th>\n",
       "      <th>data2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>b</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>b</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>b</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>a</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>a</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>a</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  key  data1  data2\n",
       "0   b      0      1\n",
       "1   b      1      1\n",
       "2   b      6      1\n",
       "3   a      2      0\n",
       "4   a      4      0\n",
       "5   a      5      0"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# you can also specify the join-column explicitly\n",
    "pd.merge(df1, df2, on='key')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can also merge with different keys:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "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>lkey</th>\n",
       "      <th>data1</th>\n",
       "      <th>rkey</th>\n",
       "      <th>data2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>b</td>\n",
       "      <td>0</td>\n",
       "      <td>b</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>b</td>\n",
       "      <td>1</td>\n",
       "      <td>b</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>b</td>\n",
       "      <td>6</td>\n",
       "      <td>b</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>a</td>\n",
       "      <td>2</td>\n",
       "      <td>a</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>a</td>\n",
       "      <td>4</td>\n",
       "      <td>a</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>a</td>\n",
       "      <td>5</td>\n",
       "      <td>a</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  lkey  data1 rkey  data2\n",
       "0    b      0    b      1\n",
       "1    b      1    b      1\n",
       "2    b      6    b      1\n",
       "3    a      2    a      0\n",
       "4    a      4    a      0\n",
       "5    a      5    a      0"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df3 = pd.DataFrame({'lkey': ['b', 'b', 'a', 'c', 'a', 'a', 'b'],\n",
    "                    'data1': range(7)})\n",
    "df4 = pd.DataFrame({'rkey': ['a', 'b', 'd'],\n",
    "                    'data2': range(3)})\n",
    "pd.merge(df3, df4, left_on='lkey', right_on='rkey')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is the outer join. Non-matches will receive NaN."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "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>key</th>\n",
       "      <th>data1</th>\n",
       "      <th>data2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>b</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>b</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>b</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>a</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>a</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>a</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>c</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>d</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  key  data1  data2\n",
       "0   b    0.0    1.0\n",
       "1   b    1.0    1.0\n",
       "2   b    6.0    1.0\n",
       "3   a    2.0    0.0\n",
       "4   a    4.0    0.0\n",
       "5   a    5.0    0.0\n",
       "6   c    3.0    NaN\n",
       "7   d    NaN    2.0"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.merge(df1, df2, how='outer')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can also join on multiple columns:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "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>key1</th>\n",
       "      <th>key2</th>\n",
       "      <th>lval</th>\n",
       "      <th>rval</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>foo</td>\n",
       "      <td>one</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>foo</td>\n",
       "      <td>one</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>foo</td>\n",
       "      <td>two</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>bar</td>\n",
       "      <td>one</td>\n",
       "      <td>3.0</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>bar</td>\n",
       "      <td>two</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  key1 key2  lval  rval\n",
       "0  foo  one   1.0   4.0\n",
       "1  foo  one   1.0   5.0\n",
       "2  foo  two   2.0   NaN\n",
       "3  bar  one   3.0   6.0\n",
       "4  bar  two   NaN   7.0"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "left = pd.DataFrame({'key1': ['foo', 'foo', 'bar'],\n",
    "                     'key2': ['one', 'two', 'one'],\n",
    "                     'lval': [1, 2, 3]})\n",
    "right = pd.DataFrame({'key1': ['foo', 'foo', 'bar', 'bar'],\n",
    "                      'key2': ['one', 'one', 'one', 'two'],\n",
    "                      'rval': [4, 5, 6, 7]})\n",
    "pd.merge(left, right, on=['key1', 'key2'], how='outer')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Another common way to tables indices is to join on indices:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "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>Ohio</th>\n",
       "      <th>Nevada</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>e</th>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Ohio  Nevada\n",
       "a   1.0     2.0\n",
       "c   3.0     4.0\n",
       "e   5.0     6.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>Missouri</th>\n",
       "      <th>Alabama</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>9.0</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d</th>\n",
       "      <td>11.0</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>e</th>\n",
       "      <td>13.0</td>\n",
       "      <td>14.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Missouri  Alabama\n",
       "b       7.0      8.0\n",
       "c       9.0     10.0\n",
       "d      11.0     12.0\n",
       "e      13.0     14.0"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>Ohio</th>\n",
       "      <th>Nevada</th>\n",
       "      <th>Missouri</th>\n",
       "      <th>Alabama</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.0</td>\n",
       "      <td>8.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11.0</td>\n",
       "      <td>12.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>e</th>\n",
       "      <td>5.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>14.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Ohio  Nevada  Missouri  Alabama\n",
       "a   1.0     2.0       NaN      NaN\n",
       "b   NaN     NaN       7.0      8.0\n",
       "c   3.0     4.0       9.0     10.0\n",
       "d   NaN     NaN      11.0     12.0\n",
       "e   5.0     6.0      13.0     14.0"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "left2 = pd.DataFrame([[1., 2.], [3., 4.], [5., 6.]],\n",
    "                     index=['a', 'c', 'e'],\n",
    "                     columns=['Ohio', 'Nevada'])\n",
    "right2 = pd.DataFrame([[7., 8.], [9., 10.], [11., 12.], [13, 14]],\n",
    "                      index=['b', 'c', 'd', 'e'],\n",
    "                      columns=['Missouri', 'Alabama'])\n",
    "display(left2)\n",
    "display(right2)\n",
    "pd.merge(left2, right2, how='outer', left_index=True, right_index=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Concatenation\n",
    "\n",
    "Another very common operation is to append dataframes by another one along its rows or columns.\n",
    "\n",
    "Recall from our `numpy` lesson:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0,  1,  2,  3],\n",
       "       [ 4,  5,  6,  7],\n",
       "       [ 8,  9, 10, 11]])"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "array([[ 0,  1,  2,  3,  0,  1,  2,  3],\n",
       "       [ 4,  5,  6,  7,  4,  5,  6,  7],\n",
       "       [ 8,  9, 10, 11,  8,  9, 10, 11]])"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "arr = np.arange(12).reshape((3, 4))\n",
    "display(arr)\n",
    "np.concatenate([arr, arr], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "s1 = pd.Series([0, 1], index=['a', 'b'])\n",
    "s2 = pd.Series([2, 3, 4], index=['c', 'd', 'e'])\n",
    "s3 = pd.Series([5, 6], index=['f', 'g'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "a    0\n",
       "b    1\n",
       "c    2\n",
       "d    3\n",
       "e    4\n",
       "f    5\n",
       "g    6\n",
       "dtype: int64"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# by default, concat works on axis=0 (row append)\n",
    "pd.concat([s1, s2, s3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>a</th>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>c</th>\n",
       "      <td>NaN</td>\n",
       "      <td>2.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d</th>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>e</th>\n",
       "      <td>NaN</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>g</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     0    1    2\n",
       "a  0.0  NaN  NaN\n",
       "b  1.0  NaN  NaN\n",
       "c  NaN  2.0  NaN\n",
       "d  NaN  3.0  NaN\n",
       "e  NaN  4.0  NaN\n",
       "f  NaN  NaN  5.0\n",
       "g  NaN  NaN  6.0"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# append in column direction\n",
    "pd.concat([s1, s2, s3], axis=1, sort=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are many more reshape operations and all kinds of data manupilations. We will use them when we need it. Read up on many of the great pandas tutorials on the net..."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Let's work with some real data now...\n",
    "\n",
    "We take the forest fire dataset to practice our skills a bit. First check the location in the linux file system and peek at the first rows."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-rw-r--r--  1 fuchs-k  IMK-IFU\\Domain Users  25478 Aug 27 15:29 ../data/uci/forestfire/forestfires.csv\n"
     ]
    }
   ],
   "source": [
    "! ls -l ../data/uci/forestfire/forestfires.csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X,Y,month,day,FFMC,DMC,DC,ISI,temp,RH,wind,rain,area\n",
      "7,5,mar,fri,86.2,26.2,94.3,5.1,8.2,51,6.7,0,0\n",
      "7,4,oct,tue,90.6,35.4,669.1,6.7,18,33,0.9,0,0\n",
      "7,4,oct,sat,90.6,43.7,686.9,6.7,14.6,33,1.3,0,0\n",
      "8,6,mar,fri,91.7,33.3,77.5,9,8.3,97,4,0.2,0\n",
      "8,6,mar,sun,89.3,51.3,102.2,9.6,11.4,99,1.8,0,0\n",
      "8,6,aug,sun,92.3,85.3,488,14.7,22.2,29,5.4,0,0\n",
      "8,6,aug,mon,92.3,88.9,495.6,8.5,24.1,27,3.1,0,0\n",
      "8,6,aug,mon,91.5,145.4,608.2,10.7,8,86,2.2,0,0\n",
      "8,6,sep,tue,91,129.5,692.6,7,13.1,63,5.4,0,0\n"
     ]
    }
   ],
   "source": [
    "! head ../data/uci/forestfire/forestfires.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So it's comma delimited, has a header row and looks pretty bog standard. No need to use any other hundreds of extra features of `read_csv()` while reading it in."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('../data/uci/forestfire/forestfires.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After reading the file always check the dimensions, data types of the columns and the first couple of rows to see they the reading worked as expected."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "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>X</th>\n",
       "      <th>Y</th>\n",
       "      <th>month</th>\n",
       "      <th>day</th>\n",
       "      <th>FFMC</th>\n",
       "      <th>DMC</th>\n",
       "      <th>DC</th>\n",
       "      <th>ISI</th>\n",
       "      <th>temp</th>\n",
       "      <th>RH</th>\n",
       "      <th>wind</th>\n",
       "      <th>rain</th>\n",
       "      <th>area</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>mar</td>\n",
       "      <td>fri</td>\n",
       "      <td>86.2</td>\n",
       "      <td>26.2</td>\n",
       "      <td>94.3</td>\n",
       "      <td>5.1</td>\n",
       "      <td>8.2</td>\n",
       "      <td>51</td>\n",
       "      <td>6.7</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>7</td>\n",
       "      <td>4</td>\n",
       "      <td>oct</td>\n",
       "      <td>tue</td>\n",
       "      <td>90.6</td>\n",
       "      <td>35.4</td>\n",
       "      <td>669.1</td>\n",
       "      <td>6.7</td>\n",
       "      <td>18.0</td>\n",
       "      <td>33</td>\n",
       "      <td>0.9</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>7</td>\n",
       "      <td>4</td>\n",
       "      <td>oct</td>\n",
       "      <td>sat</td>\n",
       "      <td>90.6</td>\n",
       "      <td>43.7</td>\n",
       "      <td>686.9</td>\n",
       "      <td>6.7</td>\n",
       "      <td>14.6</td>\n",
       "      <td>33</td>\n",
       "      <td>1.3</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>8</td>\n",
       "      <td>6</td>\n",
       "      <td>mar</td>\n",
       "      <td>fri</td>\n",
       "      <td>91.7</td>\n",
       "      <td>33.3</td>\n",
       "      <td>77.5</td>\n",
       "      <td>9.0</td>\n",
       "      <td>8.3</td>\n",
       "      <td>97</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.2</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8</td>\n",
       "      <td>6</td>\n",
       "      <td>mar</td>\n",
       "      <td>sun</td>\n",
       "      <td>89.3</td>\n",
       "      <td>51.3</td>\n",
       "      <td>102.2</td>\n",
       "      <td>9.6</td>\n",
       "      <td>11.4</td>\n",
       "      <td>99</td>\n",
       "      <td>1.8</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   X  Y month  day  FFMC   DMC     DC  ISI  temp  RH  wind  rain  area\n",
       "0  7  5   mar  fri  86.2  26.2   94.3  5.1   8.2  51   6.7   0.0   0.0\n",
       "1  7  4   oct  tue  90.6  35.4  669.1  6.7  18.0  33   0.9   0.0   0.0\n",
       "2  7  4   oct  sat  90.6  43.7  686.9  6.7  14.6  33   1.3   0.0   0.0\n",
       "3  8  6   mar  fri  91.7  33.3   77.5  9.0   8.3  97   4.0   0.2   0.0\n",
       "4  8  6   mar  sun  89.3  51.3  102.2  9.6  11.4  99   1.8   0.0   0.0"
      ]
     },
     "execution_count": 179,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The first rows of the dataframe\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(517, 13)"
      ]
     },
     "execution_count": 180,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# the shape of the dataframe\n",
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "X          int64\n",
       "Y          int64\n",
       "month     object\n",
       "day       object\n",
       "FFMC     float64\n",
       "DMC      float64\n",
       "DC       float64\n",
       "ISI      float64\n",
       "temp     float64\n",
       "RH         int64\n",
       "wind     float64\n",
       "rain     float64\n",
       "area     float64\n",
       "dtype: object"
      ]
     },
     "execution_count": 181,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# the data types\n",
    "df.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Columns 'month' and 'day' are of type object (this means 'string' or anything else pandas can not do computations with). However, they should actually be of type `categorical` as the values have to be from a defined set of values (month names, day names).\n",
    "\n",
    "So let's convert the dtype of those columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[mar, oct, oct, mar, mar, ..., aug, aug, aug, aug, nov]\n",
       "Length: 517\n",
       "Categories (12, object): [apr, aug, dec, feb, ..., may, nov, oct, sep]"
      ]
     },
     "execution_count": 182,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# column month converted to categorical dtype\n",
    "c1 = pd.Categorical(df.month)\n",
    "c1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[mar, oct, oct, mar, mar, ..., aug, aug, aug, aug, nov]\n",
       "Length: 517\n",
       "Categories (12, object): [jan, feb, mar, apr, ..., sep, oct, nov, dec]"
      ]
     },
     "execution_count": 183,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# column month converted to categorical dtype and giving a category set and order\n",
    "c2 = pd.Categorical(df.month, categories=['jan','feb','mar','apr','may','jun','jul','aug','sep','oct','nov','dec'])\n",
    "c2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [],
   "source": [
    "# assign the new column back to the dataframe\n",
    "df['month'] = c2"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The dataframe contains records of burned areas (column \"area\"). Let's quickly visualize the distribution."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.area.hist(bins=20);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Seems there are actually a lot of records without any burned area (area==0). Let's count them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "247"
      ]
     },
     "execution_count": 186,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df[df.area == 0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We want to find out when most forest fires occur. So let's filter the data. We can simply use `value_counts()` on the month column to find this out:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "aug    184\n",
       "sep    172\n",
       "mar     54\n",
       "jul     32\n",
       "feb     20\n",
       "jun     17\n",
       "oct     15\n",
       "dec      9\n",
       "apr      9\n",
       "may      2\n",
       "jan      2\n",
       "nov      1\n",
       "Name: month, dtype: int64"
      ]
     },
     "execution_count": 187,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.month.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "However, this also includes the rows with no fires. So we filter them first. We also normlize the counts and plot it in one line. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df[df.area > 0].month.value_counts(normalize=True).plot(kind='bar');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This actualy shows another power of pandas. It's very easy to chain operations and to directly visualize them. Pandas has some plotting directives incorporated, we will get to know more on that in the plotting lesson."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Working with time-series data\n",
    "\n",
    "Pandas was originally developed for business applications (like stock trading actions) and thus has very strong time-series capabilities. This is great since handling dates, times and time-series in general usually is a major pain!\n",
    "\n",
    "We now look at some raw micromet sensor data from a permanent field measurement site of the TERENO network in Fendt near the IMK-IFU institute.\n",
    "\n",
    "We one file per day containing records of a large range of senors in a 1-minute resolution."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-rwxr-xr-x  1 fuchs-k  IMK-IFU\\Domain Users  723911 Aug 27 15:29 ../data/tereno_fendt_sample/Fen_M_18_001.dat\n",
      "-rwxr-xr-x  1 fuchs-k  IMK-IFU\\Domain Users  723291 Aug 27 15:29 ../data/tereno_fendt_sample/Fen_M_18_002.dat\n",
      "-rwxr-xr-x  1 fuchs-k  IMK-IFU\\Domain Users  724284 Aug 27 15:29 ../data/tereno_fendt_sample/Fen_M_18_003.dat\n",
      "-rwxr-xr-x  1 fuchs-k  IMK-IFU\\Domain Users  723743 Aug 27 15:29 ../data/tereno_fendt_sample/Fen_M_18_004.dat\n",
      "-rwxr-xr-x  1 fuchs-k  IMK-IFU\\Domain Users  718091 Aug 27 15:29 ../data/tereno_fendt_sample/Fen_M_18_005.dat\n"
     ]
    }
   ],
   "source": [
    "# data\n",
    "! ls -l ../data/tereno_fendt_sample/*dat | head -n 5"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With ! we can directly execute bash commands from python, no need to open a seperate terminal. The files are labeled with the year (18) and the doy (001, 002, ...)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\"2018-01-01 00:00:00\",13.6,-0.053,180,199,211,0.6,0.8,1,0.6,89.1,939,0,0,0,0,0,0,1.363,1.863,2.114,2.329,2.429,-79.23,1.542,1.784,2.212,2.399,2.461,2.956,1.537,1.836,2.142,2.052,2.494,\"NAN\",0.748,0.748,0.779,0.779,0.779,0.779,0.765,0.731,0.73,0.778,0.779,\"NAN\",0.75,\"NAN\",0.757,0.793,0.78,\"NAN\",-0.015,0,0.167,-4.112,-7.482,-23.77,-81,2.24,275.4,-3.371,-57.27,0.549,-27.88,-88.5,-60.64,302.3,245.1,-8.002541,-7.806843,940,-1.93,17.77,18.43,0,-11.27593,-10.07353,-9.107799,0.02866541,0.04813046,0.04811176\n",
      "\"2018-01-01 00:01:00\",13.33,-0.053,136,150,167,0.3,0.5,0.7,0.5,88.5,939,0,0,0,0,0,0,1.343,1.872,2.183,2.319,2.43,-81,1.529,1.797,2.224,2.431,2.441,2.952,1.483,1.826,2.182,2.062,2.537,\"NAN\",0.748,0.748,0.778,0.779,0.779,0.779,0.765,0.731,0.73,0.777,0.779,\"NAN\",0.749,\"NAN\",0.757,0.793,0.779,\"NAN\",-0.012,0,0,-4.361,-8.24,-23.88,-79.06,2.246,275.4,-3.881,-55.19,0.529,-28.24,-87.3,-59.07,302.3,247.1,-7.977484,-7.806873,940,-1.972,0.987,1.81,0,-11.27593,-10.03892,-9.107799,0.02866541,0.04813046,0.04811176\n"
     ]
    }
   ],
   "source": [
    "# peek into one file to infer the file format\n",
    "! head -n 2 ../data/tereno_fendt_sample/Fen_M_18_001.dat"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we can see there's a lot going on. There's no header labeling the columns, there's NaN values, there's a timestamp in column 1 and the file is comma delimited. \n",
    "\n",
    "Luckily there exists a data file description (in Excel format no less :-/). So I exportewd the column names from this description that we can find in file colnames.csv. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TIMESTAMP,BattV_Avg,PTemp_C_Avg,Wdmin_Min,Wdavg,Wdmax_Max,Wsmin_Min,Wsavg_Avg,Wsmax_Max,airtemp_Avg,relhumidity_Avg,airpressure_Avg,Ramount,Rduration_Avg,Rintensity_Avg,Hamount,Hduration_Avg,Hintensity_Avg,T107_2_West_Avg,T107_6_West_Avg,T107_12_West_Avg,T107_25_West_Avg,T107_35_West_Avg,T107_50_West_Avg,T107_2_Mitte_Avg,T107_6_Mitte_Avg,T107_12_Mitte_Avg,T107_25_Mitte_Avg,T107_35_Mitte_Avg,T107_50_Mitte_Avg,T107_2_Ost_Avg,T107_6_Ost_Avg,T107_12_Ost_Avg,T107_25_Ost_Avg,T107_35_Ost_Avg,T107_50_Ost_Avg,VWC_2_West_Avg,VWC_6_West_Avg,VWC_12_West_Avg,VWC_25_West_Avg,VWC_35_West_Avg,VWC_50_West_Avg,VWC_2_Mitte_Avg,VWC_6_Mitte_Avg,VWC_12_Mitte_Avg,VWC_25_Mitte_Avg,VWC_35_Mitte_Avg,VWC_50_Mitte_Avg,VWC_2_Ost_Avg,VWC_6_Ost_Avg,VWC_12_Ost_Avg,VWC_25_Ost_Avg,VWC_35_Ost_Avg,VWC_50_Ost_Avg,Snow_Depth,PAR_1_Avg,PAR_2_Avg,SR_Out_Avg,SR_In_Avg,IR_Out_Avg,IR_In_Avg,NR01_TC_Avg/CNR4TC_Avg,NR01_TK_Avg/CNR4TK_Avg,NetRs_Avg,NetRl_Avg,Albedo_Avg,OutTot_Avg,InTot_Avg,NetTot_Avg,IR_OutCo_Avg,IR_InCo_Avg,H_Flux_8_Ost_neu_Avg,H_Flux_9_West_neu_Avg,BP_mbar_Avg,IR_TempC_Avg,Total_Avg,Diffuse_Avg,Sun,H_Flux_sc_9_Ost_Avg,H_Flux_sc_8_fernerOst_Avg,H_Flux_sc_8_Mitte_Avg,shf_cal(1),shf_cal(2),shf_cal(3)\n"
     ]
    }
   ],
   "source": [
    "! cat ../data/tereno_fendt_sample/colnames.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we read the colnames and store them in a list."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['TIMESTAMP',\n",
       " 'BattV_Avg',\n",
       " 'PTemp_C_Avg',\n",
       " 'Wdmin_Min',\n",
       " 'Wdavg',\n",
       " 'Wdmax_Max',\n",
       " 'Wsmin_Min',\n",
       " 'Wsavg_Avg',\n",
       " 'Wsmax_Max',\n",
       " 'airtemp_Avg',\n",
       " 'relhumidity_Avg',\n",
       " 'airpressure_Avg',\n",
       " 'Ramount',\n",
       " 'Rduration_Avg',\n",
       " 'Rintensity_Avg',\n",
       " 'Hamount',\n",
       " 'Hduration_Avg',\n",
       " 'Hintensity_Avg',\n",
       " 'T107_2_West_Avg',\n",
       " 'T107_6_West_Avg',\n",
       " 'T107_12_West_Avg',\n",
       " 'T107_25_West_Avg',\n",
       " 'T107_35_West_Avg',\n",
       " 'T107_50_West_Avg',\n",
       " 'T107_2_Mitte_Avg',\n",
       " 'T107_6_Mitte_Avg',\n",
       " 'T107_12_Mitte_Avg',\n",
       " 'T107_25_Mitte_Avg',\n",
       " 'T107_35_Mitte_Avg',\n",
       " 'T107_50_Mitte_Avg',\n",
       " 'T107_2_Ost_Avg',\n",
       " 'T107_6_Ost_Avg',\n",
       " 'T107_12_Ost_Avg',\n",
       " 'T107_25_Ost_Avg',\n",
       " 'T107_35_Ost_Avg',\n",
       " 'T107_50_Ost_Avg',\n",
       " 'VWC_2_West_Avg',\n",
       " 'VWC_6_West_Avg',\n",
       " 'VWC_12_West_Avg',\n",
       " 'VWC_25_West_Avg',\n",
       " 'VWC_35_West_Avg',\n",
       " 'VWC_50_West_Avg',\n",
       " 'VWC_2_Mitte_Avg',\n",
       " 'VWC_6_Mitte_Avg',\n",
       " 'VWC_12_Mitte_Avg',\n",
       " 'VWC_25_Mitte_Avg',\n",
       " 'VWC_35_Mitte_Avg',\n",
       " 'VWC_50_Mitte_Avg',\n",
       " 'VWC_2_Ost_Avg',\n",
       " 'VWC_6_Ost_Avg',\n",
       " 'VWC_12_Ost_Avg',\n",
       " 'VWC_25_Ost_Avg',\n",
       " 'VWC_35_Ost_Avg',\n",
       " 'VWC_50_Ost_Avg',\n",
       " 'Snow_Depth',\n",
       " 'PAR_1_Avg',\n",
       " 'PAR_2_Avg',\n",
       " 'SR_Out_Avg',\n",
       " 'SR_In_Avg',\n",
       " 'IR_Out_Avg',\n",
       " 'IR_In_Avg',\n",
       " 'NR01_TC_Avg/CNR4TC_Avg',\n",
       " 'NR01_TK_Avg/CNR4TK_Avg',\n",
       " 'NetRs_Avg',\n",
       " 'NetRl_Avg',\n",
       " 'Albedo_Avg',\n",
       " 'OutTot_Avg',\n",
       " 'InTot_Avg',\n",
       " 'NetTot_Avg',\n",
       " 'IR_OutCo_Avg',\n",
       " 'IR_InCo_Avg',\n",
       " 'H_Flux_8_Ost_neu_Avg',\n",
       " 'H_Flux_9_West_neu_Avg',\n",
       " 'BP_mbar_Avg',\n",
       " 'IR_TempC_Avg',\n",
       " 'Total_Avg',\n",
       " 'Diffuse_Avg',\n",
       " 'Sun',\n",
       " 'H_Flux_sc_9_Ost_Avg',\n",
       " 'H_Flux_sc_8_fernerOst_Avg',\n",
       " 'H_Flux_sc_8_Mitte_Avg',\n",
       " 'shf_cal(1)',\n",
       " 'shf_cal(2)',\n",
       " 'shf_cal(3)']"
      ]
     },
     "execution_count": 192,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "COLNAMES = open('../data/tereno_fendt_sample/colnames.csv').readline()[:-1].split(',')\n",
    "COLNAMES"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then we read one actual data file and assign those column names. \n",
    "\n",
    "We also specify some rules for the import:\n",
    "- there is a datetime column we want parsed into a datetime dtype called \"TIMESTAMP\"\n",
    "- we have no header in the file (data starts in the first row)\n",
    "- missing data values in the file are marked with the string \"NAN\"\n",
    "\n",
    "After reading the file we set the index of the dataframe to be the timestamp column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('../data/tereno_fendt_sample/Fen_M_18_001.dat', \n",
    "                 names=COLNAMES, \n",
    "                 parse_dates=['TIMESTAMP'],\n",
    "                 header=None, sep=',', na_values=['NAN'])\n",
    "df = df.set_index('TIMESTAMP')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As always, we check the import and some dataframe features (shape, dtypes)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "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>BattV_Avg</th>\n",
       "      <th>PTemp_C_Avg</th>\n",
       "      <th>Wdmin_Min</th>\n",
       "      <th>Wdavg</th>\n",
       "      <th>Wdmax_Max</th>\n",
       "      <th>Wsmin_Min</th>\n",
       "      <th>Wsavg_Avg</th>\n",
       "      <th>Wsmax_Max</th>\n",
       "      <th>airtemp_Avg</th>\n",
       "      <th>relhumidity_Avg</th>\n",
       "      <th>...</th>\n",
       "      <th>IR_TempC_Avg</th>\n",
       "      <th>Total_Avg</th>\n",
       "      <th>Diffuse_Avg</th>\n",
       "      <th>Sun</th>\n",
       "      <th>H_Flux_sc_9_Ost_Avg</th>\n",
       "      <th>H_Flux_sc_8_fernerOst_Avg</th>\n",
       "      <th>H_Flux_sc_8_Mitte_Avg</th>\n",
       "      <th>shf_cal(1)</th>\n",
       "      <th>shf_cal(2)</th>\n",
       "      <th>shf_cal(3)</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TIMESTAMP</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2018-01-01 00:00:00</th>\n",
       "      <td>13.60</td>\n",
       "      <td>-0.053</td>\n",
       "      <td>180</td>\n",
       "      <td>199</td>\n",
       "      <td>211</td>\n",
       "      <td>0.6</td>\n",
       "      <td>0.8</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.6</td>\n",
       "      <td>89.1</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.930</td>\n",
       "      <td>17.770</td>\n",
       "      <td>18.430</td>\n",
       "      <td>0</td>\n",
       "      <td>-11.27593</td>\n",
       "      <td>-10.07353</td>\n",
       "      <td>-9.107799</td>\n",
       "      <td>0.028665</td>\n",
       "      <td>0.04813</td>\n",
       "      <td>0.048112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-01 00:01:00</th>\n",
       "      <td>13.33</td>\n",
       "      <td>-0.053</td>\n",
       "      <td>136</td>\n",
       "      <td>150</td>\n",
       "      <td>167</td>\n",
       "      <td>0.3</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.7</td>\n",
       "      <td>0.5</td>\n",
       "      <td>88.5</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.972</td>\n",
       "      <td>0.987</td>\n",
       "      <td>1.810</td>\n",
       "      <td>0</td>\n",
       "      <td>-11.27593</td>\n",
       "      <td>-10.03892</td>\n",
       "      <td>-9.107799</td>\n",
       "      <td>0.028665</td>\n",
       "      <td>0.04813</td>\n",
       "      <td>0.048112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-01 00:02:00</th>\n",
       "      <td>13.33</td>\n",
       "      <td>-0.056</td>\n",
       "      <td>156</td>\n",
       "      <td>166</td>\n",
       "      <td>183</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.7</td>\n",
       "      <td>0.8</td>\n",
       "      <td>0.5</td>\n",
       "      <td>88.3</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.902</td>\n",
       "      <td>1.645</td>\n",
       "      <td>2.797</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.028665</td>\n",
       "      <td>0.04813</td>\n",
       "      <td>0.048112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-01 00:03:00</th>\n",
       "      <td>13.52</td>\n",
       "      <td>-0.058</td>\n",
       "      <td>162</td>\n",
       "      <td>178</td>\n",
       "      <td>187</td>\n",
       "      <td>0.6</td>\n",
       "      <td>0.7</td>\n",
       "      <td>0.8</td>\n",
       "      <td>0.5</td>\n",
       "      <td>88.5</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.874</td>\n",
       "      <td>17.930</td>\n",
       "      <td>2.139</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.028665</td>\n",
       "      <td>0.04813</td>\n",
       "      <td>0.048112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018-01-01 00:04:00</th>\n",
       "      <td>13.58</td>\n",
       "      <td>-0.053</td>\n",
       "      <td>172</td>\n",
       "      <td>180</td>\n",
       "      <td>188</td>\n",
       "      <td>0.7</td>\n",
       "      <td>0.9</td>\n",
       "      <td>1.1</td>\n",
       "      <td>0.5</td>\n",
       "      <td>88.8</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.741</td>\n",
       "      <td>17.440</td>\n",
       "      <td>18.590</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.028665</td>\n",
       "      <td>0.04813</td>\n",
       "      <td>0.048112</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 83 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     BattV_Avg  PTemp_C_Avg  Wdmin_Min  Wdavg  Wdmax_Max  \\\n",
       "TIMESTAMP                                                                  \n",
       "2018-01-01 00:00:00      13.60       -0.053        180    199        211   \n",
       "2018-01-01 00:01:00      13.33       -0.053        136    150        167   \n",
       "2018-01-01 00:02:00      13.33       -0.056        156    166        183   \n",
       "2018-01-01 00:03:00      13.52       -0.058        162    178        187   \n",
       "2018-01-01 00:04:00      13.58       -0.053        172    180        188   \n",
       "\n",
       "                     Wsmin_Min  Wsavg_Avg  Wsmax_Max  airtemp_Avg  \\\n",
       "TIMESTAMP                                                           \n",
       "2018-01-01 00:00:00        0.6        0.8        1.0          0.6   \n",
       "2018-01-01 00:01:00        0.3        0.5        0.7          0.5   \n",
       "2018-01-01 00:02:00        0.5        0.7        0.8          0.5   \n",
       "2018-01-01 00:03:00        0.6        0.7        0.8          0.5   \n",
       "2018-01-01 00:04:00        0.7        0.9        1.1          0.5   \n",
       "\n",
       "                     relhumidity_Avg  ...  IR_TempC_Avg  Total_Avg  \\\n",
       "TIMESTAMP                             ...                            \n",
       "2018-01-01 00:00:00             89.1  ...        -1.930     17.770   \n",
       "2018-01-01 00:01:00             88.5  ...        -1.972      0.987   \n",
       "2018-01-01 00:02:00             88.3  ...        -1.902      1.645   \n",
       "2018-01-01 00:03:00             88.5  ...        -1.874     17.930   \n",
       "2018-01-01 00:04:00             88.8  ...        -1.741     17.440   \n",
       "\n",
       "                     Diffuse_Avg  Sun  H_Flux_sc_9_Ost_Avg  \\\n",
       "TIMESTAMP                                                    \n",
       "2018-01-01 00:00:00       18.430    0            -11.27593   \n",
       "2018-01-01 00:01:00        1.810    0            -11.27593   \n",
       "2018-01-01 00:02:00        2.797    0                  NaN   \n",
       "2018-01-01 00:03:00        2.139    0                  NaN   \n",
       "2018-01-01 00:04:00       18.590    0                  NaN   \n",
       "\n",
       "                     H_Flux_sc_8_fernerOst_Avg  H_Flux_sc_8_Mitte_Avg  \\\n",
       "TIMESTAMP                                                               \n",
       "2018-01-01 00:00:00                  -10.07353              -9.107799   \n",
       "2018-01-01 00:01:00                  -10.03892              -9.107799   \n",
       "2018-01-01 00:02:00                        NaN                    NaN   \n",
       "2018-01-01 00:03:00                        NaN                    NaN   \n",
       "2018-01-01 00:04:00                        NaN                    NaN   \n",
       "\n",
       "                     shf_cal(1)  shf_cal(2)  shf_cal(3)  \n",
       "TIMESTAMP                                                \n",
       "2018-01-01 00:00:00    0.028665     0.04813    0.048112  \n",
       "2018-01-01 00:01:00    0.028665     0.04813    0.048112  \n",
       "2018-01-01 00:02:00    0.028665     0.04813    0.048112  \n",
       "2018-01-01 00:03:00    0.028665     0.04813    0.048112  \n",
       "2018-01-01 00:04:00    0.028665     0.04813    0.048112  \n",
       "\n",
       "[5 rows x 83 columns]"
      ]
     },
     "execution_count": 194,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1440, 83)"
      ]
     },
     "execution_count": 195,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "BattV_Avg                    float64\n",
       "PTemp_C_Avg                  float64\n",
       "Wdmin_Min                      int64\n",
       "Wdavg                          int64\n",
       "Wdmax_Max                      int64\n",
       "Wsmin_Min                    float64\n",
       "Wsavg_Avg                    float64\n",
       "Wsmax_Max                    float64\n",
       "airtemp_Avg                  float64\n",
       "relhumidity_Avg              float64\n",
       "airpressure_Avg                int64\n",
       "Ramount                      float64\n",
       "Rduration_Avg                  int64\n",
       "Rintensity_Avg               float64\n",
       "Hamount                        int64\n",
       "Hduration_Avg                  int64\n",
       "Hintensity_Avg                 int64\n",
       "T107_2_West_Avg              float64\n",
       "T107_6_West_Avg              float64\n",
       "T107_12_West_Avg             float64\n",
       "T107_25_West_Avg             float64\n",
       "T107_35_West_Avg             float64\n",
       "T107_50_West_Avg             float64\n",
       "T107_2_Mitte_Avg             float64\n",
       "T107_6_Mitte_Avg             float64\n",
       "T107_12_Mitte_Avg            float64\n",
       "T107_25_Mitte_Avg            float64\n",
       "T107_35_Mitte_Avg            float64\n",
       "T107_50_Mitte_Avg            float64\n",
       "T107_2_Ost_Avg               float64\n",
       "                              ...   \n",
       "Snow_Depth                   float64\n",
       "PAR_1_Avg                    float64\n",
       "PAR_2_Avg                    float64\n",
       "SR_Out_Avg                   float64\n",
       "SR_In_Avg                    float64\n",
       "IR_Out_Avg                   float64\n",
       "IR_In_Avg                    float64\n",
       "NR01_TC_Avg/CNR4TC_Avg       float64\n",
       "NR01_TK_Avg/CNR4TK_Avg       float64\n",
       "NetRs_Avg                    float64\n",
       "NetRl_Avg                    float64\n",
       "Albedo_Avg                   float64\n",
       "OutTot_Avg                   float64\n",
       "InTot_Avg                    float64\n",
       "NetTot_Avg                   float64\n",
       "IR_OutCo_Avg                 float64\n",
       "IR_InCo_Avg                  float64\n",
       "H_Flux_8_Ost_neu_Avg         float64\n",
       "H_Flux_9_West_neu_Avg        float64\n",
       "BP_mbar_Avg                    int64\n",
       "IR_TempC_Avg                 float64\n",
       "Total_Avg                    float64\n",
       "Diffuse_Avg                  float64\n",
       "Sun                            int64\n",
       "H_Flux_sc_9_Ost_Avg          float64\n",
       "H_Flux_sc_8_fernerOst_Avg    float64\n",
       "H_Flux_sc_8_Mitte_Avg        float64\n",
       "shf_cal(1)                   float64\n",
       "shf_cal(2)                   float64\n",
       "shf_cal(3)                   float64\n",
       "Length: 83, dtype: object"
      ]
     },
     "execution_count": 196,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note, since there are so many columns and rows jupyter shortens the print. If you want to have a look at all of them you need to modify the defaults:\n",
    "\n",
    "```\n",
    "pd.set_option('display.height', 1000)\n",
    "pd.set_option('display.max_rows', 500)\n",
    "pd.set_option('display.max_columns', 500)\n",
    "pd.set_option('display.width', 1000)\n",
    "```\n",
    "\n",
    "\n",
    "Let's look at a single column:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig,ax = plt.subplots(1,1, figsize=(14,3))\n",
    "df.airtemp_Avg.plot();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ok, now we read in six more files and combine them into one larger file covering 1 week of data.\n",
    "\n",
    "We use a small utility function to do this. This function does several things at once:\n",
    "- it reads in one or multiple daily files\n",
    "- it parses them as done before\n",
    "- it set's the timestamp column as the dataframe index\n",
    "- it concatenates the individual dataframes into a single on (axis-0 is vertically)\n",
    "- it finally sorts the index (timestamp) ans thus the dataframe in ascending order"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {},
   "outputs": [],
   "source": [
    "def read_tereno_daily_micromet(filename, COLNAMES):\n",
    "    \"\"\"read one or multiple files and return a joined dataframe\"\"\"\n",
    "    if isinstance(filename, str):\n",
    "        filename = [filename]\n",
    "    \n",
    "    fs = []\n",
    "    for f in filename:\n",
    "        df = pd.read_csv(f, \n",
    "                     names=COLNAMES, \n",
    "                     parse_dates=['TIMESTAMP'],\n",
    "                     header=None, sep=',', na_values=['NAN'])\n",
    "        df = df.set_index('TIMESTAMP')\n",
    "        fs.append(df)\n",
    "    return pd.concat(fs, axis=0).sort_index()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we read 7 daily files. We use the glob module to specify a range of files with wildcard syntax (* or [1-7])."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {},
   "outputs": [],
   "source": [
    "# let's read 7 days of data\n",
    "import glob\n",
    "\n",
    "# this gives us the files 001, 002, 003, ... 007\n",
    "filenames = glob.glob('../data/tereno_fendt_sample/Fen_M_18_00[1-7]*.dat')\n",
    "df = read_tereno_daily_micromet(filenames, COLNAMES)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10080, 83)"
      ]
     },
     "execution_count": 200,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TIMESTAMP\n",
       "2018-01-01 00:00:00    0.6\n",
       "2018-01-01 00:01:00    0.5\n",
       "2018-01-01 00:02:00    0.5\n",
       "2018-01-01 00:03:00    0.5\n",
       "2018-01-01 00:04:00    0.5\n",
       "Name: airtemp_Avg, dtype: float64"
      ]
     },
     "execution_count": 201,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# the head of the dataframe\n",
    "df.airtemp_Avg.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TIMESTAMP\n",
       "2018-01-07 23:55:00    1.6\n",
       "2018-01-07 23:56:00    1.6\n",
       "2018-01-07 23:57:00    1.6\n",
       "2018-01-07 23:58:00    1.6\n",
       "2018-01-07 23:59:00    1.6\n",
       "Name: airtemp_Avg, dtype: float64"
      ]
     },
     "execution_count": 202,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# the last rows of the dataframe\n",
    "df.airtemp_Avg.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plotting again\n",
    "fig,ax = plt.subplots(1,1, figsize=(14,3))\n",
    "df.airtemp_Avg.plot();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Notice that pandas automatically adjusts the x-axis to days (previously hours). It can do this since the index is encoded as a proper datetime dtype.\n",
    "\n",
    "## Working with datetime encoded data\n",
    "\n",
    "Queriung and indexing time and dates in pandas is kinda magic. See this example.\n",
    "\n",
    "Say we want to get all data from 3:00 to 9:00 for the 2nd of January. We can index into the dataframe datetime-index column."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# select a certain period from datetime index\n",
    "subset = df['2018-01-02 03:00':'2018-01-02 09:00']\n",
    "fig,ax = plt.subplots(1,1, figsize=(14,3))\n",
    "subset.airtemp_Avg.plot();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also select all rows of the day 2018-01-06 using the following syntax (it essentially rounds the datetime index down to daily values)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# select a single day from datetime index\n",
    "oneday_df = df['2018-01-06']\n",
    "fig,ax = plt.subplots(1,1, figsize=(14,3))\n",
    "oneday_df.airtemp_Avg.plot();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Pandas comes also with resampling tools for building different temporal resolutions.\n",
    "\n",
    "Say, we want to change our data to hourly values. We can simply do that be calling `resample()` with the appropriate time unit abbreviations (day=D, hour=H, or 15min for 15 minute intervals).\n",
    "\n",
    "Ignore the plotting command in the next cell (we do plotting later)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x216 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig,ax = plt.subplots(1,4, figsize=(16,3))\n",
    "\n",
    "one_day = df['2018-01-04']\n",
    "\n",
    "one_day.airtemp_Avg.plot(ax=ax[0], label='original', color='black')\n",
    "one_day.airtemp_Avg.resample('15min').mean().plot(ax=ax[1], label='15min', color='purple')\n",
    "one_day.airtemp_Avg.resample('1H').mean().plot(ax=ax[2], label='1H', color='blue')\n",
    "one_day.airtemp_Avg.resample('3H').mean().plot(ax=ax[3], label='3H', color='orange')\n",
    "\n",
    "for a in ax:\n",
    "    a.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also use a  sliding window (moving average) approach to resample. This is called `rolling` in pandas and works similar to resampling.\n",
    "\n",
    "Let's say we want to smooth with a sliding window of 30 data points (centered in the middle, this will lead to 14 NaN values at the beginning and at the end of the timesereis). The data resolution stays at 1 data point per minute."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# rolling window/ centered with window size = 30 datapoints\n",
    "fig,ax = plt.subplots(1,1, figsize=(12,3))\n",
    "oneday_df.airtemp_Avg.plot(ax=ax);\n",
    "oneday_df.airtemp_Avg.rolling(30, center=True).mean().plot(ax=ax);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}