mieszkania5/Lab1_zad2.ipynb at master

import pandas as pd
import numpy as np
import joblib

df_train = pd.read_csv('train/train.tsv', sep='\t', header=None)

pd.set_option('display.max_columns', 500)

df_train.head()

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25
0	309000.0	do zamieszkania	390 zł	spółdzielcze własnościowe	7113	https://www.otodom.pl/oferta/niezalezny-uklad-...	2	NaN	43.44	wtórny	4.0	blok	NaN	NaN	Niezależny Układ W Nowoczesnym Wydaniu	1	NaN	gazowe	plastikowe	NaN	NaN	NaN	cegła	Polecamy na sprzedaż dwupokojowe mieszkanie p...	NaN	telewizja kablowa, internet, meble, piwnica, g...
1	314900.0	do wykończenia	NaN	pełna własność	7392	https://www.otodom.pl/oferta/urokliwe-mieszkan...	2	NaN	42.60	pierwotny	2.0	blok	NaN	NaN	Urokliwe mieszkanie 2 pokojowe Strzeszyn	1	NaN	gazowe	plastikowe	NaN	NaN	NaN	cegła	Kameralne 2 pokojowe mieszkanie z aneksem kuc...	NaN	telewizja kablowa, internet, telefon, drzwi / ...
2	249000.0	do remontu	300 zł	pełna własność	5621	https://www.otodom.pl/oferta/mieszkanie-do-rem...	2	NaN	44.30	wtórny	4.0	blok	NaN	NaN	Mieszkanie do remontu ul. Klonowa, blisko tramwaj	2	NaN	miejskie	plastikowe	1960.0	NaN	2019-06-30	cegła	Oferta bezpośrednio od właściciela - bez pośr...	NaN	telewizja kablowa, internet, telefon, domofon ...
3	419000.0	do zamieszkania	490 zł	pełna własność	4761	https://www.otodom.pl/oferta/w-szeregowcu-4-po...	4	NaN	88	wtórny	3.0	szeregowiec	NaN	NaN	W szeregowcu 4 pokoje z garażem	1	NaN	gazowe	plastikowe	NaN	NaN	NaN	cegła	Drodzy Państwo Zapraszam do zapoznania się z ...	NaN	telewizja kablowa, internet, zmywarka, lodówka...
4	499000.0	NaN	850 zł	NaN	6481	https://www.otodom.pl/oferta/komfortowe-przest...	3	NaN	77	wtórny	16.0	blok	NaN	NaN	Komfortowe,Przestronne,3Pokoje, Armii Krajowej!!!	7	NaN	NaN	plastikowe	NaN	NaN	NaN	NaN	Biuro Immohouse ma przyjemność proponować do ...	NaN	balkon, piwnica, winda

1,2,6, 8, 9 - to bedziemy w aplikacji używać

df_train = df_train.iloc[:, [0, 1, 2, 6, 8, 9]]

df_train.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 2547 entries, 0 to 2546
Data columns (total 6 columns):
 #   Column  Non-Null Count  Dtype  
---  ------  --------------  -----  
 0   0       2547 non-null   float64
 1   1       1519 non-null   object 
 2   2       929 non-null    object 
 3   6       2547 non-null   object 
 4   8       2547 non-null   object 
 5   9       2547 non-null   object 
dtypes: float64(1), object(5)
memory usage: 119.5+ KB

df_train.rename(columns={0: "cena m", 1: "stan", 2: "czynsz", 6: "l pokoi", 8: "metraż", 9: "rynek"}, inplace=True)

df_train.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 2547 entries, 0 to 2546
Data columns (total 6 columns):
 #   Column   Non-Null Count  Dtype  
---  ------   --------------  -----  
 0   cena m   2547 non-null   float64
 1   stan     1519 non-null   object 
 2   czynsz   929 non-null    object 
 3   l pokoi  2547 non-null   object 
 4   metraż   2547 non-null   object 
 5   rynek    2547 non-null   object 
dtypes: float64(1), object(5)
memory usage: 119.5+ KB

Usunięcie kolumn, w których jest dużo pustych wartości

df_train.drop(columns = ['czynsz'], axis=1, inplace=True)

df_train.head()

	cena m	stan	l pokoi	metraż	rynek
0	309000.0	do zamieszkania	2	43.44	wtórny
1	314900.0	do wykończenia	2	42.60	pierwotny
2	249000.0	do remontu	2	44.30	wtórny
3	419000.0	do zamieszkania	4	88	wtórny
4	499000.0	NaN	3	77	wtórny

Uzupełnianie braków w danych

df_train.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 2547 entries, 0 to 2546
Data columns (total 5 columns):
 #   Column   Non-Null Count  Dtype  
---  ------   --------------  -----  
 0   cena m   2547 non-null   float64
 1   stan     1519 non-null   object 
 2   l pokoi  2547 non-null   object 
 3   metraż   2547 non-null   object 
 4   rynek    2547 non-null   object 
dtypes: float64(1), object(4)
memory usage: 99.6+ KB

df_train['stan'].fillna('brak info', inplace=True)

df_train = df_train[df_train['metraż'] != "6 909"]

df_train = df_train[df_train['l pokoi'] != "więcej niż 10"]

df_train['l pokoi'] = pd.to_numeric(df_train['l pokoi'])

df_train['metraż'] = pd.to_numeric(df_train['metraż'])

X_train = df_train.drop(columns = ["cena m"])
y_train = df_train["cena m"]

number_cols = X_train.select_dtypes(exclude=[object]).columns
object_cols = X_train.select_dtypes(include=[object]).columns

from sklearn.compose import ColumnTransformer
from sklearn.preprocessing import OneHotEncoder
from sklearn.preprocessing import StandardScaler
from sklearn.pipeline import make_pipeline

ct = ColumnTransformer([
        (number_cols, StandardScaler()),
        (object_cols, OneHotEncoder())
    ],remainder='passthrough')

from sklearn.linear_model import Ridge

scaler = StandardScaler()
encoder = OneHotEncoder()
# putting numeric columns to scaler and categorical to encoder
num_transformer = make_pipeline(scaler)
cat_transformer = make_pipeline(encoder)

# getting together our scaler and encoder with preprocessor
preprocessor = ColumnTransformer(
      transformers=[('num', num_transformer, number_cols),
                    ('cat', cat_transformer, object_cols)])

clf = Ridge(alpha=1.0)

model1 = make_pipeline(preprocessor, clf)

X_train.head()

	stan	l pokoi	metraż	rynek
0	do zamieszkania	2	43.44	wtórny
1	do wykończenia	2	42.60	pierwotny
2	do remontu	2	44.30	wtórny
3	do zamieszkania	4	88.00	wtórny
4	brak info	3	77.00	wtórny

y_train

0       309000.0
1       314900.0
2       249000.0
3       419000.0
4       499000.0
          ...   
2541    383680.0
2542    507600.0
2543    342400.0
2544    335000.0
2545    260000.0
Name: cena m, Length: 2540, dtype: float64

Model Ridge - regresja liniowa

model1.fit(X_train, y_train)

Pipeline(steps=[('columntransformer',
                 ColumnTransformer(transformers=[('num',
                                                  Pipeline(steps=[('standardscaler',
                                                                   StandardScaler())]),
                                                  Index(['l pokoi', 'metraż'], dtype='object')),
                                                 ('cat',
                                                  Pipeline(steps=[('onehotencoder',
                                                                   OneHotEncoder())]),
                                                  Index(['stan', 'rynek'], dtype='object'))])),
                ('ridge', Ridge())])

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

Pipeline

Pipeline(steps=[('columntransformer',
                 ColumnTransformer(transformers=[('num',
                                                  Pipeline(steps=[('standardscaler',
                                                                   StandardScaler())]),
                                                  Index(['l pokoi', 'metraż'], dtype='object')),
                                                 ('cat',
                                                  Pipeline(steps=[('onehotencoder',
                                                                   OneHotEncoder())]),
                                                  Index(['stan', 'rynek'], dtype='object'))])),
                ('ridge', Ridge())])

columntransformer: ColumnTransformer

ColumnTransformer(transformers=[('num',
                                 Pipeline(steps=[('standardscaler',
                                                  StandardScaler())]),
                                 Index(['l pokoi', 'metraż'], dtype='object')),
                                ('cat',
                                 Pipeline(steps=[('onehotencoder',
                                                  OneHotEncoder())]),
                                 Index(['stan', 'rynek'], dtype='object'))])

num

Index(['l pokoi', 'metraż'], dtype='object')

StandardScaler

StandardScaler()

cat

Index(['stan', 'rynek'], dtype='object')

OneHotEncoder

OneHotEncoder()

Ridge

Ridge()

df_test = pd.read_csv('dev-0/in.tsv', sep='\t', header=None)

df_test.head()

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24
0	do zamieszkania	250 zł	pełna własność	6311	https://www.otodom.pl/oferta/idealne-mieszkani...	3	NaN	59.10	wtórny	4.0	blok	NaN	NaN	idealne mieszkanie do zamieszkania od zaraz	2	NaN	miejskie	plastikowe	NaN	NaN	NaN	wielka płyta	Zamieszkaj od zaraz – idealna lokalizacja OFE...	NaN	telewizja kablowa, internet, telefon, domofon ...
1	do zamieszkania	NaN	pełna własność	7868	https://www.otodom.pl/oferta/2-pokoje-38m2-po-...	2	NaN	38.00	wtórny	12.0	blok	NaN	NaN	2 pokoje / 38m2 / po remoncie / Winogrady	4	NaN	inne	plastikowe	NaN	NaN	NaN	NaN	Na sprzedaż mieszkanie dwupokojowe na osiedlu...	NaN	balkon
2	do zamieszkania	650 zł	pełna własność	5717	https://www.otodom.pl/oferta/3-pokoje-na-jezyc...	3	NaN	63.84	wtórny	4.0	blok	NaN	NaN	3 pokoje na Jeżycach blisko Rusałki	4	NaN	miejskie	plastikowe	1958.0	NaN	NaN	cegła	!! Oferta dostępna tylko u nas !! Polecam ofe...	NaN	telewizja kablowa, internet, telefon, piwnica,...
3	do zamieszkania	359 zł	spółdzielcze własnościowe	7380	https://www.otodom.pl/oferta/uniwersytet-przyr...	4	NaN	50.00	wtórny	10.0	blok	NaN	NaN	Uniwersytet Przyrodniczy - 4 pokoje - ROI 8-10%	10	NaN	miejskie	plastikowe	1975.0	NaN	2019-06-30	wielka płyta	Oferta bezpośrednio od właściciela - bez po...	NaN	telewizja kablowa, internet, telefon, drzwi / ...
4	NaN	NaN	NaN	7373	https://www.otodom.pl/oferta/mieszkanie-blisko...	3	NaN	65.62	pierwotny	3.0	NaN	NaN	NaN	Mieszkanie blisko centrum w kameralnej okolicy.	1	NaN	NaN	NaN	2020.0	NaN	2020-01-01	NaN	MIESZKANIE 3-POKOJOWE O POW.65,62M2 Mieszkani...	NaN	garaż/miejsce parkingowe, taras, pom. użytkowe

y_test = pd.read_csv('dev-0/expected.tsv', sep='\t', header=None)

X_train

	stan	l pokoi	metraż	rynek
0	do zamieszkania	2	43.44	wtórny
1	do wykończenia	2	42.60	pierwotny
2	do remontu	2	44.30	wtórny
3	do zamieszkania	4	88.00	wtórny
4	brak info	3	77.00	wtórny
...	...	...	...	...
2541	do wykończenia	3	70.40	pierwotny
2542	do wykończenia	4	94.00	wtórny
2543	brak info	2	53.50	wtórny
2544	brak info	3	55.25	wtórny
2545	brak info	3	62.00	wtórny

2540 rows × 4 columns

df_test.rename(columns={0: "stan", 1: "czynsz", 5: "l pokoi", 7: "metraż", 8: "rynek"}, inplace=True)

df_test.columns

Index([   'stan',  'czynsz',         2,         3,         4, 'l pokoi',
               6,  'metraż',   'rynek',         9,        10,        11,
              12,        13,        14,        15,        16,        17,
              18,        19,        20,        21,        22,        23,
              24],
      dtype='object')

df_test = df_test[['stan', 'l pokoi', 'metraż',  'rynek']]

df_test.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 462 entries, 0 to 461
Data columns (total 4 columns):
 #   Column   Non-Null Count  Dtype  
---  ------   --------------  -----  
 0   stan     275 non-null    object 
 1   l pokoi  462 non-null    int64  
 2   metraż   462 non-null    float64
 3   rynek    462 non-null    object 
dtypes: float64(1), int64(1), object(2)
memory usage: 14.6+ KB

df_test['stan'].fillna('brak info', inplace=True)

df_test.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 462 entries, 0 to 461
Data columns (total 4 columns):
 #   Column   Non-Null Count  Dtype  
---  ------   --------------  -----  
 0   stan     462 non-null    object 
 1   l pokoi  462 non-null    int64  
 2   metraż   462 non-null    float64
 3   rynek    462 non-null    object 
dtypes: float64(1), int64(1), object(2)
memory usage: 14.6+ KB

df_test['metraż']

0      59.10
1      38.00
2      63.84
3      50.00
4      65.62
       ...  
457    72.78
458    51.23
459    54.16
460    90.10
461    71.90
Name: metraż, Length: 462, dtype: float64

df_test.reset_index(drop=True, inplace=True)

df_test.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 462 entries, 0 to 461
Data columns (total 4 columns):
 #   Column   Non-Null Count  Dtype  
---  ------   --------------  -----  
 0   stan     462 non-null    object 
 1   l pokoi  462 non-null    int64  
 2   metraż   462 non-null    float64
 3   rynek    462 non-null    object 
dtypes: float64(1), int64(1), object(2)
memory usage: 14.6+ KB

model1

Pipeline(steps=[('columntransformer',
                 ColumnTransformer(transformers=[('num',
                                                  Pipeline(steps=[('standardscaler',
                                                                   StandardScaler())]),
                                                  Index(['l pokoi', 'metraż'], dtype='object')),
                                                 ('cat',
                                                  Pipeline(steps=[('onehotencoder',
                                                                   OneHotEncoder())]),
                                                  Index(['stan', 'rynek'], dtype='object'))])),
                ('ridge', Ridge())])

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

Pipeline

Pipeline(steps=[('columntransformer',
                 ColumnTransformer(transformers=[('num',
                                                  Pipeline(steps=[('standardscaler',
                                                                   StandardScaler())]),
                                                  Index(['l pokoi', 'metraż'], dtype='object')),
                                                 ('cat',
                                                  Pipeline(steps=[('onehotencoder',
                                                                   OneHotEncoder())]),
                                                  Index(['stan', 'rynek'], dtype='object'))])),
                ('ridge', Ridge())])

columntransformer: ColumnTransformer

ColumnTransformer(transformers=[('num',
                                 Pipeline(steps=[('standardscaler',
                                                  StandardScaler())]),
                                 Index(['l pokoi', 'metraż'], dtype='object')),
                                ('cat',
                                 Pipeline(steps=[('onehotencoder',
                                                  OneHotEncoder())]),
                                 Index(['stan', 'rynek'], dtype='object'))])

num

Index(['l pokoi', 'metraż'], dtype='object')

StandardScaler

StandardScaler()

cat

Index(['stan', 'rynek'], dtype='object')

OneHotEncoder

OneHotEncoder()

Ridge

Ridge()

Przykład - przewidywanie cen na danych testowych

model1.predict(df_test)

array([ 407917.66889452,  285306.27863261,  444184.94702291,
        299459.05926666,  459271.83408962,  614055.95902743,
        311243.98561031,  361244.58944658,  329238.65906305,
        369470.84812889,  285350.13919824,  329238.65906305,
        420020.53940636,  399085.99304054,  368717.32379097,
        339834.65864991,  335512.74515277,  304524.88092914,
        367036.20209137,  286773.80808993,  354525.20168926,
        282085.35025252,  386303.76202814,  477660.29347794,
        407152.53644455,  490742.17175818,  529758.94244497,
        335044.40697838,  399501.21194489,  431641.29394143,
        331214.22563058,  334772.89329194,  404282.20491201,
        369968.98599038,  392538.22357404,  354476.13872373,
        531690.53563375,  428309.25045513,  298510.41733766,
        551966.54555785,  332974.0302655 ,  474293.99377424,
        247049.65613431,  383931.85143325,  546789.20856336,
        438766.28443053,  275556.44383868,  325403.7381088 ,
        319934.12593671,  277654.95413295,  294800.60980049,
        285306.27863261,  283776.01373268,  282811.35937324,
        511210.54963994,  327121.56879228,  335039.88788041,
        638797.47051695,  521922.40393947,  235305.67479634,
        238447.52001532,  465392.89368935,  272046.09327325,
        848434.43816552,  512546.27689188,  382253.07914128,
        781492.5027453 ,  531716.09929903,  442847.05008064,
        498968.43044048,  272108.82871587,  275556.44383868,
        307110.38376632,  373296.51037872,  320469.71865169,
        629040.94693472,  456886.14569235,  463535.64405524,
        356361.80264533,  387288.89243253,  249184.65490398,
        305443.0398691 ,  312129.77494706,  323562.90113092,
        286773.80808993,  338640.52949322,  265910.97279499,
        414803.86094421,  262352.30513363,  273562.29729465,
        959123.60494806,  476401.54226445,  463695.54142089,
        474293.99377424,  245933.92683679,  276309.90323886,
         95163.71086422,  410945.54601501,  377887.30507851,
        332477.49594204,  374627.25336918,  354168.19912956,
        369968.98599038,  311900.23521207,  275897.03611221,
        272224.40035237,  522497.33812211,  317135.7885512 ,
        317671.38126618,  357993.86137939,  352637.93422963,
        908358.15173798,  288551.8042999 ,  398779.94006055,
        279465.70834759,  310024.57586448,  367940.58322895,
        336039.07916333,  295296.8610478 ,  395408.55510658,
        289706.87506509,  461286.74212482,  959007.7502354 ,
        671918.16381992,  357151.93260828,  573837.44243868,
        286071.41108258,  453635.41762516,  294033.30766019,
        430106.50994353,  361819.52362922,  280070.72529348,
        820481.6050142 ,  349767.60269709,  364114.92097912,
        262468.15984629,  301183.86181458,  297937.65284535,
        285306.27863261,  393955.0865278 ,  666659.89980777,
        560179.02643777,  440551.36965458,  544315.22105819,
        460711.80794218,  281978.75424428,  380105.90610726,
        501187.31454538,  343001.78445802,  567830.35093743,
        262544.39001514,  332442.95664849,  703770.99332145,
        369470.84812889,  494101.3824478 ,  273562.29729465,
        180666.17730277,  452295.35099255,  331304.51667796,
        453635.41762516,  323108.3407589 ,  507194.68912278,
        326474.92353875,  334772.89329194,  335044.40697838,
        240054.29816025,  562052.51609502,  274404.22606576,
        375397.37008062,  253585.25133645,  311900.23521207,
        365811.99017143,  339834.65864991,  395408.55510658,
        567728.27402715,  336238.53613507,  490526.69290172,
        257086.66903124,  385858.46036054,  344082.22859239,
        533220.80053368,  295941.33661099,  220577.95997966,
        291229.76787478,  200262.6085879 ,  281213.62179431,
        338640.52949322,  267288.49428108,  439786.23720462,
        400076.14612753,  301183.86181458,  479730.95326697,
        319470.24429261,  470581.73347059,  336354.39084773,
        491317.10594082,  354168.19912956,  369551.88047185,
        435462.4370933 ,  514271.0794398 ,  331214.22563058,
        234999.90489251,  240207.32465024,  395408.55510658,
        305290.01337911,  265336.03861235,  349500.60810862,
        315911.57663126,  448585.54345538,  310063.91733215,
        414803.86094421,  521922.40393947,  768450.42872618,
        285306.27863261,  414228.92676157,  408188.43671135,
        262917.91567413,  530071.86630062,  284993.07170081,
        354168.19912956,  350852.56592942,  341174.72528252,
        174670.0129817 ,  346185.25798463,  308617.25469129,
        319206.16526408,  479496.61135786,  285306.27863261,
        338832.89745087,  238677.05975031,  300193.70872759,
        284350.94791532,  312129.77494706,  288864.94629397,
        447136.31089841,  468759.4764692 ,  311243.98561031,
        773841.17824564,  309069.24514719,  311818.91979295,
        303300.6690092 ,  249129.2915516 ,  317026.62262684,
        270578.56381593,  374826.77527865,  317632.03979852,
        218540.23357609,  336303.15819187,  364613.05884061,
        419438.51620964,  348842.0716669 ,  300494.95953345,
        245456.65579176,  632191.76778196,  342245.91071247,
        321241.94011573,  345941.94044726,  654441.17458326,
        594917.92628041,  403059.87960624,  295368.85519483,
        287144.48312671,  504057.36300177,  259834.05683704,
        356196.60189099,  671431.17127702,  221915.85692194,
        266766.67936848,  862139.34958591,  253061.54980967,
        278229.88831559,  360787.39659078,  819749.12524361,
        225938.68930354,  337453.4893506 ,  417419.08907647,
        433477.61182135,  428415.84646337,  369470.84812889,
        644948.32980378,  369778.78772305,  486728.1978552 ,
        220003.02579702,  372225.04187261,  533707.33028312,
        268312.96615313,  351337.20906469,  266766.67936848,
        433665.17760489,  291427.33823234,  681260.15179974,
        386155.25463611,  361819.52362922,  378283.64910587,
        312584.05224292,  262442.87925716,  347027.18675575,
        388599.15937804,  330653.06925032,  331214.22563058,
        392424.82162787,  581111.00288747,  510178.4226015 ,
        638681.61580429,  273562.29729465,  332934.68879784,
        370402.78487122,  251183.25797831,  454520.92388575,
        327814.99017136,  317446.36062915,  270745.3681083 ,
        568020.54920476, 1041637.53573378,  445984.0931255 ,
        246474.72195167,  353326.27035845,  456478.01580208,
        268245.71161255,  476014.4569415 ,  305276.23557673,
        590784.32443641,  441316.50210455,  265910.97279499,
        319470.24429261,  266451.36768408,  331214.22563058,
        327784.90740811,  319470.24429261,  244403.66193686,
        252780.99555724,  268449.97075841,  912139.95342218,
        430604.64780502,  536459.92048882, 1156407.40322869,
        395408.55510658,  221600.76337596,  327898.59243044,
        237907.12512623,  498752.66850787,  468835.9897142 ,
        350852.56592942,  300033.9934493 ,  232365.80188326,
        533897.52855045,  232365.80188326,  369329.42975094,
       -118069.19958188,  317722.33084589,  323601.95952243,
        369203.85354041,  268053.34365489, 1051697.65952951,
        254248.58995229,  303300.6690092 ,  273638.81053964,
        244461.9836068 ,  258111.42397945,  388332.16478956,
        429074.38290509,  536459.92048882,  800353.81940598,
        298918.26415178,  407917.66889452,  388332.16478956,
        414613.66267689, 1262607.50420783,  369329.42975094,
        319129.65201908,  334197.9591093 ,  286263.77904024,
        307993.25754312,  741174.54173005,  491064.17223088,
        298429.10191855,  292957.60313228,  315911.57663126,
        327478.85442812,  451544.27942111,  317135.7885512 ,
        724682.50318048,  509137.53807322,  361833.3014316 ,
        693810.21059336,  582168.1274389 ,  212609.37224369,
        260216.62306202,  710680.29626995,  264840.07044119,
        300877.52575844,  309268.70211893,  389929.9023685 ,
        464347.27192468,  347409.46990458,  693810.21059336,
        468363.13244184,  427977.91688601,  251373.45624563,
        427817.80138193,  352408.11141849,  728508.16543031,
        749166.7415794 ,  558648.76153784,  450512.15238267,
        571124.93957025,  592161.2797702 ,  418937.46278821,
        319398.53322173,  415378.79512685,  600960.30294481,
        263047.6131269 ,  364648.06092761,  325259.97032322,
        589476.22718124,  599471.54920288,  983410.6732152 ,
        405432.07327729,  442158.14779951,  491279.65108734,
        636692.27143438,  311900.23521207,  186643.53405465,
        454252.04268309,  348902.56302718,  335848.88089601,
        377701.62590915,  377701.62590915,  353370.41400023,
        376477.41398921,  560336.57202573,  556969.98924588,
        402990.3382257 ,  568449.1456857 ,  448323.35104102,
        517259.61509264,  688264.08517628,  606585.11129723,
       1137012.09739106,  327010.51625373,  361028.82751397,
        628428.55789859,  413554.08535898,  341098.21203752,
        401415.92968398,  455318.42593893, 1619425.93740436,
        451114.99963823,  267139.98688905,  342658.5597007 ,
        304976.87138504,  438573.91647287,  304976.87138504,
        392508.02366116,  458652.63911557,  547637.54304662,
        524873.76781497,  508805.98636568,  382751.50007893,
        366338.32418199,  645108.72838401,  532942.19783316])

df_test.columns

Index(['stan', 'l pokoi', 'metraż', 'rynek'], dtype='object')

X_train['rynek'].unique()

array(['wtórny', 'pierwotny'], dtype=object)

Saving the model to file

filename = 'ridge_model.sav'
joblib.dump(model1, filename)

['ridge_model.sav']

df_test = pd.read_csv('test-A/in.tsv', sep='\t', header=None)
df_test.rename(columns={0: "stan", 1: "czynsz", 5: "l pokoi", 7: "metraż", 8: "rynek"}, inplace=True)
df_test = df_test[['stan', 'l pokoi', 'metraż',  'rynek']] 
df_test['stan'].fillna('brak info', inplace=True)
# df_test[df_test['l pokoi'] == "więcej niż 10"] = 10
df_test.loc[df_test['l pokoi'] == "więcej niż 10", 'l pokoi'] = 10
df_test = df_test[df_test['metraż'] != "6 909"]
df_test.reset_index(drop=True, inplace=True)

df_test.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 418 entries, 0 to 417
Data columns (total 4 columns):
 #   Column   Non-Null Count  Dtype  
---  ------   --------------  -----  
 0   stan     418 non-null    object 
 1   l pokoi  418 non-null    object 
 2   metraż   418 non-null    float64
 3   rynek    418 non-null    object 
dtypes: float64(1), object(3)
memory usage: 13.2+ KB

predictions = model1.predict(df_test)

prediction = pd.DataFrame(predictions, columns=['predictions']).to_csv('test-A/out.tsv', index=False, sep="\t")

72 KiB Raw Permalink Blame History Unescape Escape

Model Ridge - regresja liniowa

72 KiB

Raw Permalink Blame History