2024-05-14 15:33:08 +02:00
{
"cells": [
{
"cell_type": "code",
2024-05-14 23:54:11 +02:00
"execution_count": 75,
2024-05-14 15:33:08 +02:00
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.tree import DecisionTreeRegressor, plot_tree, export_text\n",
"from sklearn.metrics import mean_squared_error\n",
"import matplotlib.pyplot as plt\n",
"from sklearn.preprocessing import StandardScaler\n",
"from sklearn.preprocessing import MinMaxScaler\n",
"from sklearn.decomposition import PCA"
]
},
{
"cell_type": "code",
2024-05-14 23:54:11 +02:00
"execution_count": 76,
2024-05-14 15:33:08 +02:00
"metadata": {},
"outputs": [],
"source": [
"feature_names = [\n",
" 'Powierzchnia', # 1\n",
" 'Wynagrodzenie_ogolem', # 2\n",
" 'Wynagrodzenie_w_relacji_do_sredniej', # 3\n",
" 'Dochody_podatek_lesny', # 4\n",
" 'Dochody_podatek_PCC', # 5\n",
" 'Dochody_podatek_od_dzialalnosci_gospodarczej', # 6\n",
" 'Dochody_podatek_od_nieruchomosci', # 7\n",
" 'Dochody_podatek_od_spadkow', # 8\n",
" 'Dochody_podatek_od_srodkow_transportowych', # 9\n",
" 'Dochody_podatek_rolny', # 10\n",
" 'Dochody_podatek_odrebne_ustawy', # 11\n",
" 'Dochody_razem', # 12\n",
" 'Dochody_z_majatku', # 13\n",
" 'Dochody_z_najmu_i_dzierzawy', # 14\n",
" 'Dochody_z_uslug', # 15\n",
" 'Dochody_dofinansowanie_inwestycyjne', # 16\n",
" 'Dochody_dofinansowanie_razem', # 17\n",
" 'Udzialy_w_podatkach_dochodowych_od_osob_fizycznych', # 18\n",
" 'Udzialy_w_podatkach_dochodowych_od_osob_prywatnych', # 19\n",
" 'Udzialy_w_podatkach_dochodowych_razem', # 20\n",
" 'Wplywy_z_innych_lokalnych_oplat', # 21\n",
" 'Wplywy_z_oplaty_eksploatacyjnej', # 22\n",
" 'Wplywy_z_oplaty_skarbowej', # 23\n",
" 'Wplywy_z_oplaty_targowej', # 24\n",
" 'Ludnosc_ogolem', # 25\n",
" 'Ludnosc_w_wieku_poprodukcyjnym', # 26\n",
" 'Ludnosc_w_wieku_produkcyjnym', # 27\n",
" 'Ludnosc_w_wieku_produkcyjnym_mobilnym', # 28\n",
" 'Ludnosc_w_wieku_produkcyjnym_niemobilnym', # 29\n",
" 'Ludnosc_w_wieku_przedprodukcyjnym', # 30\n",
" 'Ludnosc_mezczyzni', # 31\n",
" 'Ludnosc_mezczyzni_w_wieku_poprodukcyjnym', # 32\n",
" 'Ludnosc_mezczyzni_w_wieku_produkcyjnym', # 33\n",
" 'Ludnosc_mezczyzni_w_wieku_produkcyjnym_mobilnym', # 34\n",
" 'Ludnosc_mezczyzni_w_wieku_produkcyjnym_niemobilnym', # 35\n",
" 'Ludnosc_mezczyzni_w_wieku_przedprodukcyjnym', # 36\n",
" 'Ludnosc_kobiety', # 37\n",
" 'Ludnosc_kobiety_w_wieku_poprodukcyjnym', # 38\n",
" 'Ludnosc_kobiety_w_wieku_produkcyjnym', # 39\n",
" 'Ludnosc_kobiety_w_wieku_produkcyjnym_mobilnym', # 40\n",
" 'Ludnosc_kobiety_w_wieku_produkcyjnym_niemobilnym', # 41\n",
" 'Ludnosc_kobiety_w_wieku_przedprodukcyjnym', # 42\n",
" 'Wojewodztwo_Dolnoslaskie', # 43\n",
" 'Wojewodztwo_Kujawsko_Pomorskie', # 44\n",
" 'Wojewodztwo_Lubelskie', # 45\n",
" 'Wojewodztwo_Lubuskie', # 46\n",
" 'Wojewodztwo_Lodzkie', # 47\n",
" 'Wojewodztwo_Malopolskie', # 48\n",
" 'Wojewodztwo_Mazowieckie', # 49\n",
" 'Wojewodztwo_Opolskie', # 50\n",
" 'Wojewodztwo_Podkarpackie', # 51\n",
" 'Wojewodztwo_Podlaskie', # 52\n",
" 'Wojewodztwo_Pomorskie', # 53\n",
" 'Wojewodztwo_Slaskie', # 54\n",
" 'Wojewodztwo_Swietokrzyskie', # 55\n",
" 'Wojewodztwo_Warminsko_Mazurskie', # 56\n",
" 'Wojewodztwo_Wielkopolskie', # 57\n",
" 'Wojewodztwo_Zachodniopomorskie', # 58\n",
" 'Gestosc_zaludnienia', # 59\n",
" 'Ludnosc_na_1_km2', # 60\n",
" 'Ludnosc', # 61\n",
" 'Wskaznik_urbanizacji', # 62\n",
" 'Zmiana_liczby_ludnosci', # 63\n",
" 'Saldo_migracji_na_1000_ludnosci', # 64\n",
" 'Saldo_migracji', # 65\n",
" 'Wymeldowania_do_miast_kobiety', # 66\n",
" 'Wymeldowania_do_miast_mezczyzni', # 67\n",
" 'Wymeldowania_do_miast_ogolem', # 68\n",
" 'Wymeldowania_na_wies_kobiety', # 69\n",
" 'Wymeldowania_na_wies_mezczyzni', # 70\n",
" 'Wymeldowania_na_wies_ogolem', # 71\n",
" 'Wymeldowania_kobiety', # 72\n",
" 'Wymeldowania_mezczyzni', # 73\n",
" 'Wymeldowania_ogolem', # 74\n",
" 'Zameldowania_kobiety', # 75\n",
" 'Zameldowania_mezczyzni', # 76\n",
" 'Zameldowania_ogolem', # 77\n",
" 'Zameldowania_z_miast_kobiety', # 78\n",
" 'Zameldowania_z_miast_mezczyzni', # 79\n",
" 'Zameldowania_z_miast_ogolem', # 80\n",
" 'Zameldowania_ze_wsi_kobiety', # 81\n",
" 'Zameldowania_ze_wsi_mezczyzni', # 82\n",
" 'Zameldowania_ze_wsi_ogolem', # 83\n",
" 'Miejsca_noclegowe_caloroczne', # 84\n",
" 'Miejsca_noclegowe_ogolem', # 85\n",
" 'Obiekty_caloroczne', # 86\n",
" 'Obiekty_ogolem', # 87\n",
" 'Turysci_ogolem', # 88\n",
" 'Turysci_zagraniczni', # 89\n",
" 'Bezrobotni_do_25_roku_zycia', # 90\n",
" 'Dlugotrwale_bezrobotni', # 91\n",
" 'Bezrobotne_kobiety', # 92\n",
" 'Bezrobotni_mezczyzni', # 93\n",
" 'Bezrobotni_ogolem', # 94\n",
" 'Bezrobotni_powyzej_50_roku_zycia', # 95\n",
" 'Gmina_miejska', # 96\n",
" 'Gmina_miejsko_wiejska', # 97\n",
2024-05-14 23:54:11 +02:00
" 'Gmina_wiejska', # 98\n",
" 'Odleglosc_Warszawa', # 99\n",
" 'Odleglosc_od_centrum_decyzyjnego'] # 100\n",
2024-05-14 15:33:08 +02:00
"\n",
"all_columns = ['Kod', 'Rok'] + feature_names"
]
},
{
"cell_type": "code",
2024-05-14 23:54:11 +02:00
"execution_count": 77,
2024-05-14 15:33:08 +02:00
"metadata": {},
"outputs": [],
"source": [
"df_data = pd.read_csv(\n",
" 'dane1.csv',\n",
" encoding='ISO-8859-2',\n",
" converters={'Kod': str})"
]
},
{
"cell_type": "code",
2024-05-14 23:54:11 +02:00
"execution_count": 78,
"metadata": {},
"outputs": [],
"source": [
"df_odl = pd.read_csv(\n",
" 'gminy_centroid.csv',\n",
" encoding='ISO-8859-2',\n",
" converters={'TERYT': str})\n",
"df_odl['TERYT'] = df_odl['TERYT'].astype('str')\n",
"df_odl = df_odl[['TERYT', 'odl_Wawa', 'odl_woj']]\n",
"df_odl = df_odl.rename(columns={\n",
" 'TERYT': 'Kod',\n",
" 'odl_Wawa': 'Odleglosc_Warszawa',\n",
" 'odl_woj': 'Odleglosc_od_centrum_decyzyjnego'})"
]
},
{
"cell_type": "code",
"execution_count": 79,
"metadata": {},
"outputs": [],
"source": [
"df_data = df_data.merge(df_odl, on=['Kod'], how='left')"
]
},
{
"cell_type": "code",
"execution_count": 80,
2024-05-14 15:33:08 +02:00
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"['Program Operacyjny Infrastruktura i ?rodowisko 2014-2020'\n",
" 'Program Operacyjny Inteligentny RozwĂłj'\n",
" 'Program Operacyjny Polska Cyfrowa'\n",
" 'Program Operacyjny Wiedza Edukacja RozwĂłj'\n",
" 'Program Operacyjny Polska Wschodnia']\n"
]
}
],
"source": [
"print(df_data['Program_operacyjny'].drop_duplicates().values)"
]
},
{
"cell_type": "code",
2024-05-14 23:54:11 +02:00
"execution_count": 81,
2024-05-14 15:33:08 +02:00
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
2024-05-14 23:54:11 +02:00
"C:\\Users\\micha\\AppData\\Local\\Temp\\ipykernel_18980\\2237990575.py:27: PerformanceWarning: DataFrame is highly fragmented. This is usually the result of calling `frame.insert` many times, which has poor performance. Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
" df_data = df_data.groupby(all_columns)['Suma'].sum().reset_index()\n",
"C:\\Users\\micha\\AppData\\Local\\Temp\\ipykernel_18980\\2237990575.py:27: PerformanceWarning: DataFrame is highly fragmented. This is usually the result of calling `frame.insert` many times, which has poor performance. Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
" df_data = df_data.groupby(all_columns)['Suma'].sum().reset_index()\n",
"C:\\Users\\micha\\AppData\\Local\\Temp\\ipykernel_18980\\2237990575.py:27: PerformanceWarning: DataFrame is highly fragmented. This is usually the result of calling `frame.insert` many times, which has poor performance. Consider joining all columns at once using pd.concat(axis=1) instead. To get a de-fragmented frame, use `newframe = frame.copy()`\n",
2024-05-14 15:33:08 +02:00
" df_data = df_data.groupby(all_columns)['Suma'].sum().reset_index()\n"
]
}
],
"source": [
"# Wybór programu operacyjnego...\n",
"df_data = df_data.loc[df_data['Program_operacyjny'] == 'Program Operacyjny Infrastruktura i ?rodowisko 2014-2020'].reset_index(drop=True)\n",
2024-05-14 23:54:11 +02:00
"# df_data = df_data.loc[df_data['Program_operacyjny'] == 'Program Operacyjny Inteligentny RozwĂłj'].reset_index(drop=True)\n",
"# df_data = df_data.loc[df_data['Program_operacyjny'] == 'Program Operacyjny Polska Cyfrowa'].reset_index(drop=True)\n",
"# df_data = df_data.loc[df_data['Program_operacyjny'] == 'Program Operacyjny Wiedza Edukacja RozwĂłj'].reset_index(drop=True)\n",
"# df_data = df_data.loc[df_data['Program_operacyjny'] == 'Program Operacyjny Polska Wschodnia'].reset_index(drop=True)\n",
2024-05-14 15:33:08 +02:00
"\n",
"# Uzupełnienie brakujących danych...\n",
"df_data[[\n",
" 'Miejsca_noclegowe_caloroczne',\n",
" 'Miejsca_noclegowe_ogolem',\n",
" 'Obiekty_caloroczne',\n",
" 'Obiekty_ogolem',\n",
" 'Turysci_ogolem',\n",
" 'Turysci_zagraniczni']] = df_data[[\n",
" 'Miejsca_noclegowe_caloroczne',\n",
" 'Miejsca_noclegowe_ogolem',\n",
" 'Obiekty_caloroczne',\n",
" 'Obiekty_ogolem',\n",
" 'Turysci_ogolem',\n",
" 'Turysci_zagraniczni']].fillna(0)\n",
"\n",
"# Usunięcie niepotrzebnych rzędów...\n",
"df_data.dropna(inplace=True)\n",
"df_data = df_data[df_data['Suma'] > 0]\n",
"\n",
"df_data = df_data.groupby(all_columns)['Suma'].sum().reset_index()\n",
"\n",
"# ...\n",
2024-05-14 23:54:11 +02:00
"df_data['Suma'] = df_data['Suma'] / df_data['Ludnosc']"
2024-05-14 15:33:08 +02:00
]
},
{
"cell_type": "code",
2024-05-14 23:54:11 +02:00
"execution_count": 82,
2024-05-14 15:33:08 +02:00
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
2024-05-14 23:54:11 +02:00
"Mean Squared Error: 24357114749293.223\n"
2024-05-14 15:33:08 +02:00
]
}
],
"source": [
"X = df_data[feature_names]\n",
"\n",
"# scaler = MinMaxScaler()\n",
"# X = scaler.fit_transform(X)\n",
"\n",
"y = df_data['Suma']\n",
"\n",
"color_column = (df_data['Gestosc_zaludnienia'] > 1.5).astype(int)\n",
"\n",
"X_train, X_test, y_train, y_test, color_column_train, color_column_test = train_test_split(X, y, color_column, test_size=0.2, random_state=99)\n",
"\n",
"model = DecisionTreeRegressor(random_state=99)\n",
"model.fit(X_train, y_train)\n",
"\n",
"y_pred = model.predict(X_test)\n",
"mse = mean_squared_error(y_test, y_pred)\n",
"print('Mean Squared Error:', mse)"
]
},
{
"cell_type": "code",
2024-05-14 23:54:11 +02:00
"execution_count": 83,
2024-05-14 15:33:08 +02:00
"metadata": {},
"outputs": [
{
"data": {
2024-05-14 23:54:11 +02:00
"image/png": "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
2024-05-14 15:33:08 +02:00
"text/plain": [
"<Figure size 640x480 with 1 Axes>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plt.scatter(y_test, y_pred, alpha=0.5, c=color_column_test, cmap='viridis')\n",
"plt.xlabel('Actual')\n",
"plt.ylabel('Predicted')\n",
"plt.title('Actual vs Predicted')\n",
"\n",
2024-05-14 23:54:11 +02:00
"# plt.xlim(0, max(max(y_test), max(y_pred)))\n",
"# plt.ylim(0, max(max(y_test), max(y_pred)))\n",
2024-05-14 15:33:08 +02:00
"\n",
"# plt.xlim(0, 3*10**7)\n",
"# plt.ylim(0, 3*10**7)\n",
"\n",
"# plt.xlim(0, 3*10**6)\n",
"# plt.ylim(0, 3*10**6)\n",
"\n",
"# plt.xlim(0, 3*10**5)\n",
"# plt.ylim(0, 3*10**5)\n",
"\n",
"plt.show()"
]
},
{
"cell_type": "code",
2024-05-14 23:54:11 +02:00
"execution_count": 84,
2024-05-14 15:33:08 +02:00
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
2024-05-14 23:54:11 +02:00
"Gmina_wiejska 0.135239\n",
"Wojewodztwo_Dolnoslaskie 0.089233\n",
"Dochody_podatek_rolny 0.066759\n",
"Saldo_migracji_na_1000_ludnosci 0.059790\n",
"Zmiana_liczby_ludnosci 0.051100\n",
"Wojewodztwo_Warminsko_Mazurskie 0.047088\n",
"Wojewodztwo_Pomorskie 0.043228\n",
"Wojewodztwo_Opolskie 0.019902\n",
"Odleglosc_Warszawa 0.018391\n",
"Saldo_migracji 0.010109\n",
"Wojewodztwo_Mazowieckie 0.008000\n",
"Turysci_ogolem 0.007569\n",
"Turysci_zagraniczni 0.005412\n",
"Wojewodztwo_Podkarpackie -0.001333\n",
"Odleglosc_od_centrum_decyzyjnego -0.002845\n",
"Wplywy_z_oplaty_eksploatacyjnej -0.007789\n",
"Wojewodztwo_Swietokrzyskie -0.008502\n",
"Wojewodztwo_Zachodniopomorskie -0.009058\n",
"Powierzchnia -0.011699\n",
"Wojewodztwo_Slaskie -0.013926\n",
"Wojewodztwo_Lubelskie -0.014264\n",
"Obiekty_ogolem -0.015279\n",
"Wojewodztwo_Lubuskie -0.016251\n",
"Miejsca_noclegowe_ogolem -0.018215\n",
"Wojewodztwo_Podlaskie -0.019896\n",
"Dochody_podatek_PCC -0.020018\n",
"Dochody_podatek_od_spadkow -0.020339\n",
"Udzialy_w_podatkach_dochodowych_od_osob_prywatnych -0.020362\n",
2024-05-14 15:33:08 +02:00
"Name: Suma, dtype: float64\n"
]
}
],
"source": [
"correlation_matrix = df_data.corr()\n",
"print(correlation_matrix['Suma'].sort_values(ascending=False)[1:29])"
]
},
{
"cell_type": "code",
2024-05-14 23:54:11 +02:00
"execution_count": 85,
2024-05-14 15:33:08 +02:00
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
2024-05-14 23:54:11 +02:00
"|--- Wplywy_z_oplaty_eksploatacyjnej <= -7751.83\n",
"| |--- value: [132625880.92]\n",
"|--- Wplywy_z_oplaty_eksploatacyjnej > -7751.83\n",
"| |--- Wplywy_z_oplaty_skarbowej <= 4494.45\n",
"| | |--- Wplywy_z_oplaty_skarbowej <= 4164.20\n",
"| | | |--- Gestosc_zaludnienia <= 0.04\n",
"| | | | |--- Odleglosc_Warszawa <= 366.00\n",
"| | | | | |--- Ludnosc_kobiety_w_wieku_przedprodukcyjnym <= 160.50\n",
"| | | | | | |--- value: [4988446.69]\n",
"| | | | | |--- Ludnosc_kobiety_w_wieku_przedprodukcyjnym > 160.50\n",
"| | | | | | |--- Ludnosc_w_wieku_produkcyjnym <= 1453.50\n",
"| | | | | | | |--- value: [8029322.26]\n",
"| | | | | | |--- Ludnosc_w_wieku_produkcyjnym > 1453.50\n",
"| | | | | | | |--- value: [7792324.25]\n",
"| | | | |--- Odleglosc_Warszawa > 366.00\n",
"| | | | | |--- Dochody_podatek_lesny <= 283601.20\n",
"| | | | | | |--- value: [223720.88]\n",
"| | | | | |--- Dochody_podatek_lesny > 283601.20\n",
"| | | | | | |--- value: [147649.40]\n",
"| | | |--- Gestosc_zaludnienia > 0.04\n",
"| | | | |--- value: [40325441.22]\n",
"| | |--- Wplywy_z_oplaty_skarbowej > 4164.20\n",
"| | | |--- value: [106444841.13]\n",
"| |--- Wplywy_z_oplaty_skarbowej > 4494.45\n",
"| | |--- Wplywy_z_oplaty_skarbowej <= 12330.00\n",
"| | | |--- Dochody_podatek_lesny <= 1523.00\n",
"| | | | |--- value: [50040444.36]\n",
"| | | |--- Dochody_podatek_lesny > 1523.00\n",
"| | | | |--- Dochody_z_najmu_i_dzierzawy <= 339230.50\n",
"| | | | | |--- Bezrobotni_do_25_roku_zycia <= 41.25\n",
"| | | | | | |--- Dochody_podatek_od_srodkow_transportowych <= 136790.07\n",
"| | | | | | | |--- Zameldowania_ze_wsi_kobiety <= 5.50\n",
"| | | | | | | | |--- Dochody_razem <= 5335600.25\n",
"| | | | | | | | | |--- Wojewodztwo_Podkarpackie <= 0.50\n",
"| | | | | | | | | | |--- Dochody_z_majatku <= 98626.36\n",
"| | | | | | | | | | | |--- truncated branch of depth 3\n",
"| | | | | | | | | | |--- Dochody_z_majatku > 98626.36\n",
"| | | | | | | | | | | |--- truncated branch of depth 5\n",
"| | | | | | | | | |--- Wojewodztwo_Podkarpackie > 0.50\n",
"| | | | | | | | | | |--- Wynagrodzenie_w_relacji_do_sredniej <= 81.20\n",
"| | | | | | | | | | | |--- value: [1029930.53]\n",
"| | | | | | | | | | |--- Wynagrodzenie_w_relacji_do_sredniej > 81.20\n",
"| | | | | | | | | | | |--- value: [674809.87]\n",
"| | | | | | | | |--- Dochody_razem > 5335600.25\n",
"| | | | | | | | | |--- Wplywy_z_oplaty_skarbowej <= 10500.50\n",
"| | | | | | | | | | |--- Udzialy_w_podatkach_dochodowych_od_osob_fizycznych <= 2211601.00\n",
"| | | | | | | | | | | |--- value: [2852332.52]\n",
"| | | | | | | | | | |--- Udzialy_w_podatkach_dochodowych_od_osob_fizycznych > 2211601.00\n",
"| | | | | | | | | | | |--- value: [3787927.89]\n",
"| | | | | | | | | |--- Wplywy_z_oplaty_skarbowej > 10500.50\n",
"| | | | | | | | | | |--- Udzialy_w_podatkach_dochodowych_razem <= 1245524.19\n",
"| | | | | | | | | | | |--- value: [2301238.08]\n",
"| | | | | | | | | | |--- Udzialy_w_podatkach_dochodowych_razem > 1245524.19\n",
"| | | | | | | | | | | |--- value: [1142691.81]\n",
"| | | | | | | |--- Zameldowania_ze_wsi_kobiety > 5.50\n",
"| | | | | | | | |--- Dochody_podatek_lesny <= 16527.95\n",
"| | | | | | | | | |--- Dlugotrwale_bezrobotni <= 97.25\n",
"| | | | | | | | | | |--- Saldo_migracji_na_1000_ludnosci <= -1.20\n",
"| | | | | | | | | | | |--- value: [214308.91]\n",
"| | | | | | | | | | |--- Saldo_migracji_na_1000_ludnosci > -1.20\n",
"| | | | | | | | | | | |--- value: [248413.49]\n",
"| | | | | | | | | |--- Dlugotrwale_bezrobotni > 97.25\n",
"| | | | | | | | | | |--- Zameldowania_ze_wsi_ogolem <= 19.50\n",
"| | | | | | | | | | | |--- value: [886572.54]\n",
"| | | | | | | | | | |--- Zameldowania_ze_wsi_ogolem > 19.50\n",
"| | | | | | | | | | | |--- value: [1011703.86]\n",
"| | | | | | | | |--- Dochody_podatek_lesny > 16527.95\n",
"| | | | | | | | | |--- Zameldowania_ze_wsi_ogolem <= 9.00\n",
"| | | | | | | | | | |--- value: [621811.99]\n",
"| | | | | | | | | |--- Zameldowania_ze_wsi_ogolem > 9.00\n",
"| | | | | | | | | | |--- Bezrobotni_do_25_roku_zycia <= 13.75\n",
"| | | | | | | | | | | |--- truncated branch of depth 3\n",
"| | | | | | | | | | |--- Bezrobotni_do_25_roku_zycia > 13.75\n",
"| | | | | | | | | | | |--- truncated branch of depth 8\n",
"| | | | | | |--- Dochody_podatek_od_srodkow_transportowych > 136790.07\n",
"| | | | | | | |--- Wplywy_z_oplaty_skarbowej <= 9809.35\n",
"| | | | | | | | |--- Zameldowania_z_miast_ogolem <= 10.00\n",
"| | | | | | | | | |--- value: [1080165.70]\n",
"| | | | | | | | |--- Zameldowania_z_miast_ogolem > 10.00\n",
"| | | | | | | | | |--- Ludnosc_kobiety_w_wieku_przedprodukcyjnym <= 481.00\n",
"| | | | | | | | | | |--- Odleglosc_Warszawa <= 211.00\n",
"| | | | | | | | | | | |--- value: [3043529.06]\n",
"| | | | | | | | | | |--- Odleglosc_Warszawa > 211.00\n",
2024-05-14 15:33:08 +02:00
"| | | | | | | | | | | |--- truncated branch of depth 2\n",
2024-05-14 23:54:11 +02:00
"| | | | | | | | | |--- Ludnosc_kobiety_w_wieku_przedprodukcyjnym > 481.00\n",
"| | | | | | | | | | |--- value: [4041277.35]\n",
"| | | | | | | |--- Wplywy_z_oplaty_skarbowej > 9809.35\n",
"| | | | | | | | |--- Zameldowania_z_miast_kobiety <= 7.50\n",
"| | | | | | | | | |--- Wymeldowania_do_miast_kobiety <= 10.00\n",
"| | | | | | | | | | |--- value: [46813.47]\n",
"| | | | | | | | | |--- Wymeldowania_do_miast_kobiety > 10.00\n",
"| | | | | | | | | | |--- value: [24882.08]\n",
"| | | | | | | | |--- Zameldowania_z_miast_kobiety > 7.50\n",
"| | | | | | | | | |--- value: [325058.08]\n",
"| | | | | |--- Bezrobotni_do_25_roku_zycia > 41.25\n",
"| | | | | | |--- Dochody_podatek_od_srodkow_transportowych <= 117939.96\n",
"| | | | | | | |--- Bezrobotne_kobiety <= 86.25\n",
"| | | | | | | | |--- value: [24414621.52]\n",
"| | | | | | | |--- Bezrobotne_kobiety > 86.25\n",
"| | | | | | | | |--- Wplywy_z_oplaty_skarbowej <= 12039.50\n",
"| | | | | | | | | |--- Dochody_z_majatku <= 214076.80\n",
"| | | | | | | | | | |--- Gestosc_zaludnienia <= 0.05\n",
2024-05-14 15:33:08 +02:00
"| | | | | | | | | | | |--- truncated branch of depth 5\n",
2024-05-14 23:54:11 +02:00
"| | | | | | | | | | |--- Gestosc_zaludnienia > 0.05\n",
"| | | | | | | | | | | |--- truncated branch of depth 2\n",
"| | | | | | | | | |--- Dochody_z_majatku > 214076.80\n",
"| | | | | | | | | | |--- Ludnosc_w_wieku_przedprodukcyjnym <= 949.50\n",
"| | | | | | | | | | | |--- truncated branch of depth 2\n",
"| | | | | | | | | | |--- Ludnosc_w_wieku_przedprodukcyjnym > 949.50\n",
2024-05-14 15:33:08 +02:00
"| | | | | | | | | | | |--- truncated branch of depth 3\n",
2024-05-14 23:54:11 +02:00
"| | | | | | | | |--- Wplywy_z_oplaty_skarbowej > 12039.50\n",
"| | | | | | | | | |--- Dochody_razem <= 3644725.12\n",
"| | | | | | | | | | |--- value: [5361057.90]\n",
"| | | | | | | | | |--- Dochody_razem > 3644725.12\n",
"| | | | | | | | | | |--- value: [18697761.43]\n",
"| | | | | | |--- Dochody_podatek_od_srodkow_transportowych > 117939.96\n",
"| | | | | | | |--- Dochody_podatek_od_spadkow <= 6327.20\n",
"| | | | | | | | |--- Wymeldowania_do_miast_ogolem <= 31.00\n",
"| | | | | | | | | |--- value: [6925131.43]\n",
"| | | | | | | | |--- Wymeldowania_do_miast_ogolem > 31.00\n",
"| | | | | | | | | |--- value: [5923600.26]\n",
"| | | | | | | |--- Dochody_podatek_od_spadkow > 6327.20\n",
"| | | | | | | | |--- Wplywy_z_oplaty_skarbowej <= 9480.40\n",
"| | | | | | | | | |--- value: [16479146.65]\n",
"| | | | | | | | |--- Wplywy_z_oplaty_skarbowej > 9480.40\n",
"| | | | | | | | | |--- Ludnosc_mezczyzni <= 2635.00\n",
"| | | | | | | | | | |--- value: [19699871.32]\n",
"| | | | | | | | | |--- Ludnosc_mezczyzni > 2635.00\n",
"| | | | | | | | | | |--- Ludnosc_w_wieku_produkcyjnym_mobilnym <= 2553.00\n",
"| | | | | | | | | | | |--- value: [22040295.43]\n",
"| | | | | | | | | | |--- Ludnosc_w_wieku_produkcyjnym_mobilnym > 2553.00\n",
"| | | | | | | | | | | |--- value: [21309790.37]\n",
"| | | | |--- Dochody_z_najmu_i_dzierzawy > 339230.50\n",
"| | | | | |--- Wojewodztwo_Dolnoslaskie <= 0.50\n",
"| | | | | | |--- Miejsca_noclegowe_caloroczne <= 660.50\n",
"| | | | | | | |--- value: [2072547.09]\n",
"| | | | | | |--- Miejsca_noclegowe_caloroczne > 660.50\n",
"| | | | | | | |--- value: [22537878.79]\n",
"| | | | | |--- Wojewodztwo_Dolnoslaskie > 0.50\n",
"| | | | | | |--- value: [64551082.93]\n",
"| | |--- Wplywy_z_oplaty_skarbowej > 12330.00\n",
"| | | |--- Wymeldowania_na_wies_ogolem <= 40.50\n",
"| | | | |--- Dochody_podatek_lesny <= 19.50\n",
"| | | | | |--- Udzialy_w_podatkach_dochodowych_razem <= 2733850.38\n",
"| | | | | | |--- value: [70149861.07]\n",
"| | | | | |--- Udzialy_w_podatkach_dochodowych_razem > 2733850.38\n",
"| | | | | | |--- value: [3008745.87]\n",
"| | | | |--- Dochody_podatek_lesny > 19.50\n",
"| | | | | |--- Saldo_migracji_na_1000_ludnosci <= 20.00\n",
"| | | | | | |--- Wplywy_z_oplaty_targowej <= 1082589.50\n",
"| | | | | | | |--- Wojewodztwo_Opolskie <= 0.50\n",
"| | | | | | | | |--- Turysci_zagraniczni <= 463.50\n",
"| | | | | | | | | |--- Gestosc_zaludnienia <= 0.21\n",
"| | | | | | | | | | |--- Bezrobotni_do_25_roku_zycia <= 136.00\n",
"| | | | | | | | | | | |--- truncated branch of depth 44\n",
"| | | | | | | | | | |--- Bezrobotni_do_25_roku_zycia > 136.00\n",
"| | | | | | | | | | | |--- truncated branch of depth 7\n",
"| | | | | | | | | |--- Gestosc_zaludnienia > 0.21\n",
"| | | | | | | | | | |--- Ludnosc_na_1_km2 <= 225.60\n",
"| | | | | | | | | | | |--- truncated branch of depth 2\n",
"| | | | | | | | | | |--- Ludnosc_na_1_km2 > 225.60\n",
"| | | | | | | | | | | |--- truncated branch of depth 15\n",
"| | | | | | | | |--- Turysci_zagraniczni > 463.50\n",
"| | | | | | | | | |--- Odleglosc_od_centrum_decyzyjnego <= 37.50\n",
"| | | | | | | | | | |--- Obiekty_ogolem <= 2.00\n",
"| | | | | | | | | | | |--- value: [3322716.84]\n",
"| | | | | | | | | | |--- Obiekty_ogolem > 2.00\n",
2024-05-14 15:33:08 +02:00
"| | | | | | | | | | | |--- truncated branch of depth 4\n",
2024-05-14 23:54:11 +02:00
"| | | | | | | | | |--- Odleglosc_od_centrum_decyzyjnego > 37.50\n",
"| | | | | | | | | | |--- Dochody_razem <= 28685472.00\n",
2024-05-14 15:33:08 +02:00
"| | | | | | | | | | | |--- truncated branch of depth 6\n",
2024-05-14 23:54:11 +02:00
"| | | | | | | | | | |--- Dochody_razem > 28685472.00\n",
2024-05-14 15:33:08 +02:00
"| | | | | | | | | | | |--- truncated branch of depth 2\n",
2024-05-14 23:54:11 +02:00
"| | | | | | | |--- Wojewodztwo_Opolskie > 0.50\n",
"| | | | | | | | |--- Wplywy_z_oplaty_skarbowej <= 16267.00\n",
"| | | | | | | | | |--- value: [29647883.87]\n",
"| | | | | | | | |--- Wplywy_z_oplaty_skarbowej > 16267.00\n",
"| | | | | | | | | |--- Wynagrodzenie_ogolem <= 4434.15\n",
"| | | | | | | | | | |--- Wymeldowania_na_wies_ogolem <= 36.00\n",
"| | | | | | | | | | | |--- truncated branch of depth 5\n",
"| | | | | | | | | | |--- Wymeldowania_na_wies_ogolem > 36.00\n",
"| | | | | | | | | | | |--- truncated branch of depth 2\n",
"| | | | | | | | | |--- Wynagrodzenie_ogolem > 4434.15\n",
"| | | | | | | | | | |--- value: [11839160.68]\n",
"| | | | | | |--- Wplywy_z_oplaty_targowej > 1082589.50\n",
"| | | | | | | |--- value: [16775284.13]\n",
"| | | | | |--- Saldo_migracji_na_1000_ludnosci > 20.00\n",
"| | | | | | |--- Gestosc_zaludnienia <= 0.12\n",
"| | | | | | | |--- value: [68031871.39]\n",
"| | | | | | |--- Gestosc_zaludnienia > 0.12\n",
"| | | | | | | |--- Wymeldowania_na_wies_mezczyzni <= 14.50\n",
"| | | | | | | | |--- value: [464820.26]\n",
"| | | | | | | |--- Wymeldowania_na_wies_mezczyzni > 14.50\n",
"| | | | | | | | |--- Dochody_dofinansowanie_inwestycyjne <= 137616.73\n",
"| | | | | | | | | |--- Miejsca_noclegowe_ogolem <= 678.50\n",
"| | | | | | | | | | |--- value: [17985.39]\n",
"| | | | | | | | | |--- Miejsca_noclegowe_ogolem > 678.50\n",
"| | | | | | | | | | |--- value: [6263.24]\n",
"| | | | | | | | |--- Dochody_dofinansowanie_inwestycyjne > 137616.73\n",
"| | | | | | | | | |--- value: [121229.11]\n",
"| | | |--- Wymeldowania_na_wies_ogolem > 40.50\n",
"| | | | |--- Dochody_podatek_rolny <= 3770956.12\n",
"| | | | | |--- Turysci_ogolem <= 13.50\n",
"| | | | | | |--- Wynagrodzenie_w_relacji_do_sredniej <= 165.65\n",
"| | | | | | | |--- Ludnosc_w_wieku_poprodukcyjnym <= 5203.00\n",
"| | | | | | | | |--- Powierzchnia <= 65.50\n",
"| | | | | | | | | |--- Wplywy_z_oplaty_eksploatacyjnej <= 40867.03\n",
"| | | | | | | | | | |--- Wymeldowania_na_wies_kobiety <= 19.50\n",
"| | | | | | | | | | | |--- value: [6666654.80]\n",
"| | | | | | | | | | |--- Wymeldowania_na_wies_kobiety > 19.50\n",
"| | | | | | | | | | | |--- truncated branch of depth 22\n",
"| | | | | | | | | |--- Wplywy_z_oplaty_eksploatacyjnej > 40867.03\n",
"| | | | | | | | | | |--- Zameldowania_ze_wsi_kobiety <= 43.00\n",
2024-05-14 15:33:08 +02:00
"| | | | | | | | | | | |--- truncated branch of depth 3\n",
2024-05-14 23:54:11 +02:00
"| | | | | | | | | | |--- Zameldowania_ze_wsi_kobiety > 43.00\n",
"| | | | | | | | | | | |--- value: [11683460.24]\n",
"| | | | | | | | |--- Powierzchnia > 65.50\n",
"| | | | | | | | | |--- Wymeldowania_mezczyzni <= 29.50\n",
"| | | | | | | | | | |--- Wynagrodzenie_w_relacji_do_sredniej <= 93.20\n",
"| | | | | | | | | | | |--- truncated branch of depth 9\n",
"| | | | | | | | | | |--- Wynagrodzenie_w_relacji_do_sredniej > 93.20\n",
"| | | | | | | | | | | |--- value: [4982057.07]\n",
"| | | | | | | | | |--- Wymeldowania_mezczyzni > 29.50\n",
"| | | | | | | | | | |--- Ludnosc_mezczyzni_w_wieku_poprodukcyjnym <= 378.00\n",
"| | | | | | | | | | | |--- truncated branch of depth 7\n",
"| | | | | | | | | | |--- Ludnosc_mezczyzni_w_wieku_poprodukcyjnym > 378.00\n",
"| | | | | | | | | | | |--- truncated branch of depth 33\n",
"| | | | | | | |--- Ludnosc_w_wieku_poprodukcyjnym > 5203.00\n",
"| | | | | | | | |--- Gestosc_zaludnienia <= 3.92\n",
"| | | | | | | | | |--- Dochody_podatek_od_spadkow <= 45949.58\n",
"| | | | | | | | | | |--- Zameldowania_ze_wsi_kobiety <= 63.50\n",
"| | | | | | | | | | | |--- value: [18388.43]\n",
"| | | | | | | | | | |--- Zameldowania_ze_wsi_kobiety > 63.50\n",
"| | | | | | | | | | | |--- truncated branch of depth 3\n",
"| | | | | | | | | |--- Dochody_podatek_od_spadkow > 45949.58\n",
"| | | | | | | | | | |--- Wynagrodzenie_ogolem <= 3913.81\n",
"| | | | | | | | | | | |--- truncated branch of depth 20\n",
"| | | | | | | | | | |--- Wynagrodzenie_ogolem > 3913.81\n",
"| | | | | | | | | | | |--- truncated branch of depth 27\n",
"| | | | | | | | |--- Gestosc_zaludnienia > 3.92\n",
"| | | | | | | | | |--- value: [2622969.18]\n",
"| | | | | | |--- Wynagrodzenie_w_relacji_do_sredniej > 165.65\n",
"| | | | | | | |--- value: [8790189.40]\n",
"| | | | | |--- Turysci_ogolem > 13.50\n",
"| | | | | | |--- Udzialy_w_podatkach_dochodowych_od_osob_prywatnych <= 75710.72\n",
"| | | | | | | |--- Ludnosc_mezczyzni_w_wieku_produkcyjnym <= 5607.50\n",
"| | | | | | | | |--- Zmiana_liczby_ludnosci <= -0.40\n",
"| | | | | | | | | |--- Dochody_podatek_rolny <= 1097191.50\n",
"| | | | | | | | | | |--- Bezrobotni_mezczyzni <= 481.25\n",
"| | | | | | | | | | | |--- truncated branch of depth 3\n",
"| | | | | | | | | | |--- Bezrobotni_mezczyzni > 481.25\n",
"| | | | | | | | | | | |--- value: [2808492.26]\n",
"| | | | | | | | | |--- Dochody_podatek_rolny > 1097191.50\n",
"| | | | | | | | | | |--- Dochody_z_majatku <= 1013837.62\n",
2024-05-14 15:33:08 +02:00
"| | | | | | | | | | | |--- truncated branch of depth 2\n",
2024-05-14 23:54:11 +02:00
"| | | | | | | | | | |--- Dochody_z_majatku > 1013837.62\n",
"| | | | | | | | | | | |--- value: [1853128.66]\n",
"| | | | | | | | |--- Zmiana_liczby_ludnosci > -0.40\n",
"| | | | | | | | | |--- Zameldowania_ogolem <= 213.50\n",
"| | | | | | | | | | |--- Bezrobotni_do_25_roku_zycia <= 120.50\n",
"| | | | | | | | | | | |--- truncated branch of depth 4\n",
"| | | | | | | | | | |--- Bezrobotni_do_25_roku_zycia > 120.50\n",
"| | | | | | | | | | | |--- truncated branch of depth 5\n",
"| | | | | | | | | |--- Zameldowania_ogolem > 213.50\n",
"| | | | | | | | | | |--- value: [3185548.45]\n",
"| | | | | | | |--- Ludnosc_mezczyzni_w_wieku_produkcyjnym > 5607.50\n",
"| | | | | | | | |--- Udzialy_w_podatkach_dochodowych_razem <= 6892421.00\n",
"| | | | | | | | | |--- value: [9530029.04]\n",
"| | | | | | | | |--- Udzialy_w_podatkach_dochodowych_razem > 6892421.00\n",
"| | | | | | | | | |--- value: [14823401.88]\n",
"| | | | | | |--- Udzialy_w_podatkach_dochodowych_od_osob_prywatnych > 75710.72\n",
"| | | | | | | |--- Wojewodztwo_Opolskie <= 0.50\n",
"| | | | | | | | |--- Wplywy_z_innych_lokalnych_oplat <= 574419.50\n",
"| | | | | | | | | |--- Zmiana_liczby_ludnosci <= -6.70\n",
"| | | | | | | | | | |--- value: [13463381.79]\n",
"| | | | | | | | | |--- Zmiana_liczby_ludnosci > -6.70\n",
"| | | | | | | | | | |--- Wplywy_z_innych_lokalnych_oplat <= 532411.44\n",
"| | | | | | | | | | | |--- truncated branch of depth 11\n",
"| | | | | | | | | | |--- Wplywy_z_innych_lokalnych_oplat > 532411.44\n",
"| | | | | | | | | | | |--- value: [8471241.04]\n",
"| | | | | | | | |--- Wplywy_z_innych_lokalnych_oplat > 574419.50\n",
"| | | | | | | | | |--- Dochody_dofinansowanie_inwestycyjne <= 5859619.00\n",
"| | | | | | | | | | |--- Wynagrodzenie_ogolem <= 2693.92\n",
"| | | | | | | | | | | |--- value: [4769144.63]\n",
"| | | | | | | | | | |--- Wynagrodzenie_ogolem > 2693.92\n",
"| | | | | | | | | | | |--- truncated branch of depth 31\n",
"| | | | | | | | | |--- Dochody_dofinansowanie_inwestycyjne > 5859619.00\n",
"| | | | | | | | | | |--- Zameldowania_z_miast_ogolem <= 205.00\n",
"| | | | | | | | | | | |--- value: [12127521.41]\n",
"| | | | | | | | | | |--- Zameldowania_z_miast_ogolem > 205.00\n",
"| | | | | | | | | | | |--- truncated branch of depth 6\n",
"| | | | | | | |--- Wojewodztwo_Opolskie > 0.50\n",
"| | | | | | | | |--- Odleglosc_Warszawa <= 240.50\n",
"| | | | | | | | | |--- value: [12432814.85]\n",
"| | | | | | | | |--- Odleglosc_Warszawa > 240.50\n",
"| | | | | | | | | |--- Wynagrodzenie_ogolem <= 3246.40\n",
"| | | | | | | | | | |--- Powierzchnia <= 190.50\n",
"| | | | | | | | | | | |--- value: [3308256.35]\n",
"| | | | | | | | | | |--- Powierzchnia > 190.50\n",
"| | | | | | | | | | | |--- value: [4817436.62]\n",
"| | | | | | | | | |--- Wynagrodzenie_ogolem > 3246.40\n",
"| | | | | | | | | | |--- Turysci_ogolem <= 35482.50\n",
"| | | | | | | | | | | |--- truncated branch of depth 3\n",
"| | | | | | | | | | |--- Turysci_ogolem > 35482.50\n",
"| | | | | | | | | | | |--- value: [2447933.22]\n",
"| | | | |--- Dochody_podatek_rolny > 3770956.12\n",
"| | | | | |--- value: [9110050.20]\n",
2024-05-14 15:33:08 +02:00
"\n"
]
}
],
"source": [
"print(export_text(model, feature_names=feature_names))"
]
},
{
"cell_type": "code",
2024-05-14 23:54:11 +02:00
"execution_count": 86,
2024-05-14 15:33:08 +02:00
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
2024-05-14 23:54:11 +02:00
"0.3143 — Wplywy_z_oplaty_eksploatacyjnej\n",
"0.2451 — Wplywy_z_oplaty_skarbowej\n",
"0.0879 — Gestosc_zaludnienia\n",
"0.0826 — Dochody_podatek_lesny\n",
"0.0430 — Dochody_z_najmu_i_dzierzawy\n",
"0.0413 — Udzialy_w_podatkach_dochodowych_razem\n",
"0.0330 — Wojewodztwo_Dolnoslaskie\n",
"0.0209 — Ludnosc_na_1_km2\n",
"0.0137 — Saldo_migracji_na_1000_ludnosci\n",
"0.0126 — Bezrobotni_do_25_roku_zycia\n",
"0.0106 — Dochody_podatek_od_srodkow_transportowych\n",
"0.0085 — Bezrobotne_kobiety\n",
"0.0081 — Wymeldowania_na_wies_ogolem\n",
"0.0052 — Odleglosc_od_centrum_decyzyjnego\n",
"0.0046 — Dochody_podatek_od_spadkow\n",
"0.0044 — Wplywy_z_oplaty_targowej\n",
"0.0043 — Wojewodztwo_Opolskie\n",
"0.0042 — Ludnosc_kobiety_w_wieku_przedprodukcyjnym\n",
"0.0040 — Dochody_podatek_od_dzialalnosci_gospodarczej\n",
"0.0039 — Ludnosc_mezczyzni_w_wieku_produkcyjnym\n",
"0.0038 — Miejsca_noclegowe_caloroczne\n",
"0.0037 — Odleglosc_Warszawa\n",
"0.0035 — Dochody_razem\n",
"0.0030 — Zmiana_liczby_ludnosci\n",
"0.0028 — Wynagrodzenie_ogolem\n",
"0.0026 — Dochody_podatek_rolny\n",
"0.0025 — Turysci_zagraniczni\n",
"0.0022 — Zameldowania_z_miast_ogolem\n",
"0.0020 — Wplywy_z_innych_lokalnych_oplat\n",
"0.0018 — Wynagrodzenie_w_relacji_do_sredniej\n",
"0.0015 — Udzialy_w_podatkach_dochodowych_od_osob_prywatnych\n",
"0.0015 — Turysci_ogolem\n",
"0.0014 — Powierzchnia\n",
"0.0013 — Obiekty_ogolem\n",
"0.0012 — Zameldowania_ze_wsi_kobiety\n",
"0.0012 — Zameldowania_mezczyzni\n",
"0.0010 — Dochody_z_majatku\n",
"0.0007 — Wymeldowania_na_wies_kobiety\n",
"0.0007 — Dlugotrwale_bezrobotni\n",
"0.0007 — Bezrobotni_ogolem\n",
"0.0006 — Ludnosc_w_wieku_poprodukcyjnym\n",
"0.0006 — Dochody_z_uslug\n",
"0.0005 — Wymeldowania_kobiety\n",
"0.0005 — Saldo_migracji\n",
"0.0005 — Wymeldowania_na_wies_mezczyzni\n",
"0.0004 — Obiekty_caloroczne\n",
"0.0004 — Wymeldowania_do_miast_kobiety\n",
"0.0004 — Wymeldowania_mezczyzni\n",
"0.0004 — Dochody_dofinansowanie_inwestycyjne\n",
"0.0004 — Ludnosc_kobiety_w_wieku_produkcyjnym\n",
"0.0003 — Zameldowania_z_miast_mezczyzni\n",
"0.0003 — Ludnosc_mezczyzni_w_wieku_poprodukcyjnym\n",
"0.0003 — Zameldowania_ogolem\n",
"0.0003 — Dochody_podatek_od_nieruchomosci\n",
"0.0003 — Zameldowania_z_miast_kobiety\n",
"0.0003 — Ludnosc_w_wieku_produkcyjnym_mobilnym\n",
"0.0002 — Dochody_podatek_odrebne_ustawy\n",
"0.0002 — Dochody_podatek_PCC\n",
"0.0001 — Ludnosc_w_wieku_produkcyjnym_niemobilnym\n",
"0.0001 — Zameldowania_ze_wsi_ogolem\n",
"0.0001 — Wymeldowania_ogolem\n",
"0.0001 — Wymeldowania_do_miast_mezczyzni\n",
"0.0001 — Wojewodztwo_Lubuskie\n",
"0.0001 — Wojewodztwo_Slaskie\n",
"0.0001 — Bezrobotni_powyzej_50_roku_zycia\n",
"0.0001 — Ludnosc_mezczyzni\n",
"0.0001 — Ludnosc_kobiety_w_wieku_poprodukcyjnym\n",
"0.0001 — Ludnosc_kobiety_w_wieku_produkcyjnym_mobilnym\n",
"0.0001 — Dochody_dofinansowanie_razem\n",
"0.0001 — Zameldowania_kobiety\n",
"0.0001 — Ludnosc_mezczyzni_w_wieku_produkcyjnym_mobilnym\n",
"0.0001 — Bezrobotni_mezczyzni\n",
"0.0001 — Ludnosc_w_wieku_przedprodukcyjnym\n",
"0.0001 — Gmina_miejsko_wiejska\n",
"0.0001 — Ludnosc_mezczyzni_w_wieku_produkcyjnym_niemobilnym\n",
"0.0001 — Wymeldowania_do_miast_ogolem\n",
"0.0000 — Wskaznik_urbanizacji\n",
"0.0000 — Udzialy_w_podatkach_dochodowych_od_osob_fizycznych\n",
"0.0000 — Zameldowania_ze_wsi_mezczyzni\n",
"0.0000 — Wojewodztwo_Lubelskie\n",
"0.0000 — Ludnosc_kobiety_w_wieku_produkcyjnym_niemobilnym\n",
"0.0000 — Wojewodztwo_Pomorskie\n",
"0.0000 — Ludnosc_ogolem\n",
"0.0000 — Miejsca_noclegowe_ogolem\n",
"0.0000 — Wojewodztwo_Swietokrzyskie\n",
"0.0000 — Wojewodztwo_Malopolskie\n",
"0.0000 — Ludnosc\n",
"0.0000 — Wojewodztwo_Podkarpackie\n",
"0.0000 — Ludnosc_kobiety\n",
"0.0000 — Wojewodztwo_Lodzkie\n",
"0.0000 — Wojewodztwo_Podlaskie\n",
"0.0000 — Ludnosc_mezczyzni_w_wieku_przedprodukcyjnym\n",
"0.0000 — Ludnosc_w_wieku_produkcyjnym\n",
"0.0000 — Wojewodztwo_Warminsko_Mazurskie\n",
"0.0000 — Wojewodztwo_Mazowieckie\n",
"0.0000 — Gmina_wiejska\n",
"0.0000 — Gmina_miejska\n",
"0.0000 — Wojewodztwo_Zachodniopomorskie\n",
"0.0000 — Wojewodztwo_Kujawsko_Pomorskie\n",
"0.0000 — Wojewodztwo_Wielkopolskie\n"
2024-05-14 15:33:08 +02:00
]
}
],
"source": [
"feature_importance = dict(zip(feature_names, model.feature_importances_))\n",
"for feature, importance in sorted(feature_importance.items(), key=lambda x: x[1], reverse=True):\n",
2024-05-14 23:54:11 +02:00
" print(f'{importance:.4f} \\u2014 {feature}')"
2024-05-14 15:33:08 +02:00
]
}
],
"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": "3.11.3"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
