{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "290ea5a6",
   "metadata": {},
   "source": [
    "# Unterteilen und Kategorisieren von Daten\n",
    "\n",
    "Kontinuierliche Daten werden häufig in Bereiche unterteilt oder auf andere Weise für die Analyse gruppiert."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "391c3176",
   "metadata": {},
   "source": [
    "Angenommen, ihr habt Daten über eine Gruppe von Personen in einer Studie, die ihr in diskrete Altersgruppen einteilen möchtet. Hierfür generieren wir uns einen Dataframe mit 250 Einträgen zwischen `0` und `99`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "deec7725",
   "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>Age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>83</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>245</th>\n",
       "      <td>41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>246</th>\n",
       "      <td>95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>247</th>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>248</th>\n",
       "      <td>74</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>249</th>\n",
       "      <td>23</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>250 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Age\n",
       "0     77\n",
       "1     33\n",
       "2     50\n",
       "3     83\n",
       "4     49\n",
       "..   ...\n",
       "245   41\n",
       "246   95\n",
       "247   97\n",
       "248   74\n",
       "249   23\n",
       "\n",
       "[250 rows x 1 columns]"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "\n",
    "ages = np.random.randint(0, 99, 250)\n",
    "df = pd.DataFrame({\"Age\": ages})\n",
    "\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2c991e35",
   "metadata": {},
   "source": [
    "Anschließend bietet uns pandas mit [pandas.cut](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.cut.html) eine einfache Möglichkeit, die Ergebnisse in zehn Bereiche aufzuteilen. Um nur ganze Jahre zu erhalten, setzen wir zusätzlich `precision=0`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "e9d30508",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(68.0, 78.0], (29.0, 39.0], (48.0, 58.0], (78.0, 87.0], (48.0, 58.0], ..., (39.0, 48.0], (87.0, 97.0], (87.0, 97.0], (68.0, 78.0], (19.0, 29.0]]\n",
       "Length: 250\n",
       "Categories (10, interval[float64, right]): [(-0.1, 10.0] < (10.0, 19.0] < (19.0, 29.0] < (29.0, 39.0] ... (58.0, 68.0] < (68.0, 78.0] < (78.0, 87.0] < (87.0, 97.0]]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cats = pd.cut(ages, 10, precision=0)\n",
    "\n",
    "cats"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ca81953c",
   "metadata": {},
   "source": [
    "Mit [pandas.Categorical.categories](https://pandas.pydata.org/docs/reference/api/pandas.Categorical.categories.html) könnt ihr euch die Kategorien anzeigen lassen:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "1b8fb7b8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "IntervalIndex([(-0.1, 10.0], (10.0, 19.0], (19.0, 29.0], (29.0, 39.0],\n",
       "               (39.0, 48.0], (48.0, 58.0], (58.0, 68.0], (68.0, 78.0],\n",
       "               (78.0, 87.0], (87.0, 97.0]],\n",
       "              dtype='interval[float64, right]')"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cats.categories"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ed9275b5",
   "metadata": {},
   "source": [
    "…oder auch nur eine einzelne Kategorie:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "ab486178",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Interval(-0.1, 10.0, closed='right')"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cats.categories[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "077d8eeb",
   "metadata": {},
   "source": [
    "Mit [pandas.Categorical.codes](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.Categorical.codes.html) könnt ihr euch ein Array anzeigen lassen, in dem für jeden Wert die zugehörige Kategorie angezeigt wird:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "27faf8a0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([7, 3, 5, 8, 5, 3, 7, 5, 2, 4, 9, 4, 0, 5, 4, 8, 4, 3, 8, 9, 0, 6,\n",
       "       8, 6, 4, 0, 2, 0, 4, 2, 4, 8, 3, 6, 1, 5, 0, 7, 9, 8, 2, 6, 0, 1,\n",
       "       4, 0, 0, 8, 4, 7, 7, 6, 9, 5, 6, 7, 5, 8, 6, 1, 3, 8, 8, 0, 6, 9,\n",
       "       2, 3, 7, 1, 6, 1, 7, 4, 1, 3, 4, 7, 5, 0, 5, 9, 7, 8, 1, 9, 9, 6,\n",
       "       5, 2, 0, 8, 9, 8, 9, 8, 9, 6, 3, 6, 3, 4, 6, 9, 7, 4, 3, 9, 5, 9,\n",
       "       2, 2, 8, 4, 4, 0, 3, 9, 8, 6, 6, 9, 0, 6, 3, 1, 8, 6, 5, 4, 8, 1,\n",
       "       2, 4, 9, 9, 0, 6, 6, 8, 5, 8, 6, 2, 2, 9, 2, 8, 9, 7, 6, 6, 2, 1,\n",
       "       5, 7, 7, 0, 1, 8, 4, 9, 9, 4, 7, 8, 7, 2, 3, 5, 4, 8, 3, 6, 2, 9,\n",
       "       2, 9, 3, 5, 1, 2, 8, 6, 2, 0, 7, 3, 2, 1, 8, 1, 3, 2, 6, 6, 6, 5,\n",
       "       9, 3, 8, 8, 9, 3, 9, 8, 3, 7, 2, 9, 6, 6, 1, 9, 8, 4, 8, 0, 6, 8,\n",
       "       4, 2, 1, 9, 8, 1, 6, 4, 5, 7, 7, 8, 1, 3, 4, 6, 2, 1, 9, 3, 9, 3,\n",
       "       3, 9, 9, 4, 9, 9, 7, 2], dtype=int8)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cats.codes"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5629393d",
   "metadata": {},
   "source": [
    "Mit `value_counts` können wir uns nun anschauen, wie sich die Anzahl auf die einzelnen Bereiche verteilt:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b9656f7f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(87.0, 97.0]    36\n",
       "(78.0, 87.0]    34\n",
       "(58.0, 68.0]    32\n",
       "(39.0, 48.0]    25\n",
       "(19.0, 29.0]    24\n",
       "(29.0, 39.0]    24\n",
       "(68.0, 78.0]    21\n",
       "(10.0, 19.0]    19\n",
       "(48.0, 58.0]    18\n",
       "(-0.1, 10.0]    17\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.Series(cats).value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a3574d82",
   "metadata": {},
   "source": [
    "Auffalend ist, dass die Altersbereiche nicht gleich viele Jahre enthalten, sondern mit `20.0, 29.0` und `69.0, 78.0` zwei Bereiche nur 9 Jahre umfassen. Dies hängt damit zusammen, dass der Altersumfang nur von `0`bis `98` reicht:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "cfefc1cd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Age    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "d2f2e234",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Age    97\n",
       "dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.max()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b5ca6122",
   "metadata": {},
   "source": [
    "Mit [pandas.qcut](https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.qcut.html) wird die Menge hingegen in Bereiche unterteilt, die annähernd gleich groß sind:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "1c7cba64",
   "metadata": {},
   "outputs": [],
   "source": [
    "cats = pd.qcut(ages, 10, precision=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "92f7c0a6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-1.0, 14.0]    27\n",
       "(14.0, 26.0]    27\n",
       "(58.0, 65.0]    26\n",
       "(75.0, 84.0]    26\n",
       "(84.0, 90.0]    26\n",
       "(26.0, 34.0]    25\n",
       "(65.0, 75.0]    24\n",
       "(34.0, 45.0]    23\n",
       "(45.0, 58.0]    23\n",
       "(90.0, 97.0]    23\n",
       "Name: count, dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.Series(cats).value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "13fa613d",
   "metadata": {},
   "source": [
    "Wollen wir gewährleisten, dass jede Altersgruppe tatsächlich genau zehn Jahre umfasst, können wir dies mit [pandas.Categorical](https://pandas.pydata.org/docs/reference/api/pandas.Categorical.html) direkt angeben:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "fa011ab7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['0 - 9', '10 - 19', '100 - 109', '20 - 29', '30 - 39', '40 - 49',\n",
       "       '50 - 59', '60 - 69', '70 - 79', '80 - 89', '90 - 99'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "age_groups = [\"{0} - {1}\".format(i, i + 9) for i in range(0, 109, 10)]\n",
    "cats = pd.Categorical(age_groups)\n",
    "\n",
    "cats.categories"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c620f5a",
   "metadata": {},
   "source": [
    "Für die Gruppierung wird nun [pandas.cut](https://pandas.pydata.org/docs/reference/api/pandas.cut.html) verwendet:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "00c6b05b",
   "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>Age</th>\n",
       "      <th>Age group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>77</td>\n",
       "      <td>70 - 79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>33</td>\n",
       "      <td>30 - 39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>50</td>\n",
       "      <td>50 - 59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>83</td>\n",
       "      <td>80 - 89</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>49</td>\n",
       "      <td>40 - 49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>245</th>\n",
       "      <td>41</td>\n",
       "      <td>40 - 49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>246</th>\n",
       "      <td>95</td>\n",
       "      <td>90 - 99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>247</th>\n",
       "      <td>97</td>\n",
       "      <td>90 - 99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>248</th>\n",
       "      <td>74</td>\n",
       "      <td>70 - 79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>249</th>\n",
       "      <td>23</td>\n",
       "      <td>20 - 29</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>250 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Age Age group\n",
       "0     77   70 - 79\n",
       "1     33   30 - 39\n",
       "2     50   50 - 59\n",
       "3     83   80 - 89\n",
       "4     49   40 - 49\n",
       "..   ...       ...\n",
       "245   41   40 - 49\n",
       "246   95   90 - 99\n",
       "247   97   90 - 99\n",
       "248   74   70 - 79\n",
       "249   23   20 - 29\n",
       "\n",
       "[250 rows x 2 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[\"Age group\"] = pd.cut(df.Age, range(0, 111, 10), right=False, labels=cats)\n",
    "\n",
    "df"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.13 Kernel",
   "language": "python",
   "name": "python313"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.13.0"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {},
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}