48 KiB
48 KiB
import pandas as pd
from statistics import mean,median
import re
import numpy as np
Wczytanie datasetów
train_dataset = pd.read_csv("./train/train.tsv", sep = "\t", header=None)Data exploration
train_dataset.head(1)| 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | ... | 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 | ... | NaN | gazowe | plastikowe | NaN | NaN | NaN | cegła | Polecamy na sprzedaż dwupokojowe mieszkanie p... | NaN | telewizja kablowa, internet, meble, piwnica, g... |
1 rows × 26 columns
Wczytywanie danych testowych i preprocessing jak na treningu
# Preprocessing column 15:
print("COLUMN 15:")
# Count the occurrence of unique values in column before preprocessing:
print("Value counts before changes:\n",train_dataset[15].value_counts())
# Replace string to int or NaN:
train_dataset[15] = train_dataset[15].replace({"parter": 0, "suterena": -1, "> 10": 11, "poddasze": np.nan})
train_dataset[15] = train_dataset[15].apply(float)
# Fill Nans with median:
train_dataset[15].fillna(train_dataset[15].median(), inplace=True)
train_dataset[15]= train_dataset[15].apply(int)
# Count the occurrence of unique values in column after preprocessing:
print("Value counts after changes:\n",train_dataset[15].value_counts())
# Preprocessing column 8:
print("COLUMN 8:")
# Replace strings containing space to NaN:
train_dataset[8] = train_dataset[8].replace(' ', np.nan, regex=True)
# Fill Nans with median:
train_dataset[8] = train_dataset[8].apply(float)
train_dataset[8].fillna(train_dataset[8].median(), inplace=True)
print(train_dataset[8])
# Preprocessing column 6:
print("COLUMN 6:")
# Count the occurrence of unique values in column before preprocessing:
print("Value counts before changes:\n",train_dataset[6].value_counts())
# Change string to 10:
train_dataset[6] = train_dataset[6].replace({"więcej niż 10": 10})
train_dataset[6] = train_dataset[6].apply(int)
# Count the occurrence of unique values in column after preprocessing:
print("Value counts after changes:\n",train_dataset[6].value_counts())
train_dataset[10].fillna(train_dataset[10].median(), inplace=True)
train_dataset[10] = train_dataset[10].apply(float)
train_dataset = train_dataset[[0,6,8,10,15]]
COLUMN 15:
Value counts before changes:
15
1 569
2 527
0 452
4 357
3 321
5 117
6 51
7 42
8 32
10 29
11 24
9 21
-1 5
Name: count, dtype: int64
Value counts after changes:
15
1 569
2 527
0 452
4 357
3 321
5 117
6 51
7 42
8 32
10 29
11 24
9 21
-1 5
Name: count, dtype: int64
COLUMN 8:
0 43.44
1 42.60
2 44.30
3 88.00
4 77.00
...
2542 94.00
2543 53.50
2544 55.25
2545 62.00
2546 392.00
Name: 8, Length: 2547, dtype: float64
COLUMN 6:
Value counts before changes:
6
2 1014
3 878
4 293
1 271
5 64
6 13
7 7
10 6
9 1
Name: count, dtype: int64
Value counts after changes:
6
2 1014
3 878
4 293
1 271
5 64
6 13
7 7
10 6
9 1
Name: count, dtype: int64
test_dataset = pd.read_csv("./dev-0/in.tsv", sep= "\t", header=None)
# Preprocessing column 15:
print("COLUMN 15:")
# Count the occurrence of unique values in column before preprocessing:
print("Value counts before changes:\n",test_dataset[14].value_counts())
# Replace string to int or NaN:
test_dataset[14] = test_dataset[14].replace({"parter": 0, "suterena": -1, "> 10": 11, "poddasze": np.nan})
test_dataset[14] = test_dataset[14].apply(float)
# Fill Nans with median:
test_dataset[14].fillna(test_dataset[14].median(), inplace=True)
test_dataset[14]= test_dataset[14].apply(int)
# Count the occurrence of unique values in column after preprocessing:
print("Value counts after changes:\n",test_dataset[14].value_counts())
# Preprocessing column 8:
print("COLUMN 8:")
# Replace strings containing space to NaN:
test_dataset[7] = test_dataset[7].replace(' ', np.nan, regex=True)
# Fill Nans with median:
test_dataset[7] = test_dataset[7].apply(float)
test_dataset[7].fillna(test_dataset[7].median(), inplace=True)
print(test_dataset[7])
# Preprocessing column 6:
print("COLUMN 6:")
# Count the occurrence of unique values in column before preprocessing:
print("Value counts before changes:\n",test_dataset[5].value_counts())
# Change string to 10:
test_dataset[5] = test_dataset[5].replace({"więcej niż 10": 10})
test_dataset[5] = test_dataset[5].apply(int)
# Count the occurrence of unique values in column after preprocessing:
print("Value counts after changes:\n",test_dataset[5].value_counts())
test_dataset[9].fillna(test_dataset[9].median(), inplace=True)
test_dataset[9] = test_dataset[9].apply(float)
test_dataset = test_dataset[[5,7,9,14]]
COLUMN 15:
Value counts before changes:
14
1 108
2 89
0 82
4 65
3 54
5 22
6 12
7 9
11 9
10 5
8 3
-1 2
9 2
Name: count, dtype: int64
Value counts after changes:
14
1 108
2 89
0 82
4 65
3 54
5 22
6 12
7 9
11 9
10 5
8 3
-1 2
9 2
Name: count, dtype: int64
COLUMN 8:
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: 7, Length: 462, dtype: float64
COLUMN 6:
Value counts before changes:
5
2 196
3 152
1 51
4 50
5 9
6 4
Name: count, dtype: int64
Value counts after changes:
5
2 196
3 152
1 51
4 50
5 9
6 4
Name: count, dtype: int64
test_dataset.head(1)| 5 | 7 | 9 | 14 | |
|---|---|---|---|---|
| 0 | 3 | 59.1 | 4.0 | 2 |
Model
from sklearn.linear_model import LinearRegression
from sklearn.preprocessing import StandardScalerX_train = train_dataset.drop(0,axis=1)
y_train = train_dataset[[0]]
scaler = StandardScaler()
trans_data = scaler.fit_transform(X)X_test = test_datasetreg = LinearRegression()reg.fit(X_train, y_train)
results = reg.predict(X_test)import pickle
from sklearn.metrics import r2_score
# pickle.dump(reg, open("model.pkl", "wb"))resultsarray([[ 394901.20434554],
[ 293271.41755997],
[ 432666.21541136],
[ 295330.00173591],
[ 444001.60173013],
[ 595102.64364947],
[ 306562.53056792],
[ 346367.98664224],
[ 320708.28590668],
[ 355678.48101873],
[ 272087.02326267],
[ 315111.17058773],
[ 408743.59976314],
[ 392452.80148004],
[ 351776.80580328],
[ 325340.80253875],
[ 324415.74401286],
[ 286605.64798109],
[ 354409.52248178],
[ 273703.59297418],
[ 343402.57630023],
[ 285271.79788568],
[ 370401.6837088 ],
[ 482522.18182752],
[ 394119.81457774],
[ 474686.42204592],
[ 627962.51992074],
[ 334221.5006669 ],
[ 397338.23464708],
[ 414336.53457982],
[ 335485.00759125],
[ 351363.03195993],
[ 384755.53002533],
[ 374013.34228146],
[ 384755.53002533],
[ 355882.69461662],
[ 521392.20674093],
[ 425546.58946533],
[ 294138.97474644],
[ 539668.46177031],
[ 340107.2565533 ],
[ 467566.06735236],
[ 228315.74093461],
[ 373516.14139746],
[ 572962.98245529],
[ 425442.35590324],
[ 264688.02449027],
[ 321384.60839985],
[ 313697.70958017],
[ 257339.46910406],
[ 285491.92892354],
[ 265315.27967261],
[ 269349.61545595],
[ 370207.10222578],
[ 505152.41437514],
[ 326640.0334956 ],
[ 361868.73382815],
[ 641520.72645455],
[ 513506.74409331],
[ 225524.54295198],
[ 237226.80467502],
[ 453176.39203834],
[ 261995.60845714],
[ 955187.2509814 ],
[ 492991.37526251],
[ 374938.40080734],
[ 774416.69022809],
[ 523696.94084834],
[ 434831.26310559],
[ 489623.47044873],
[ 280423.33071801],
[ 264688.02449027],
[ 280814.31555288],
[ 359688.46533354],
[ 314210.82790415],
[ 622185.54247246],
[ 448643.38610329],
[ 561068.71341085],
[ 331068.80586855],
[ 387502.56994691],
[ 251024.38630808],
[ 295949.29222845],
[ 311115.63444521],
[ 307926.65308324],
[ 268121.19700892],
[ 329516.02282157],
[ 260145.38644037],
[ 402080.9057926 ],
[ 247029.12134696],
[ 293241.9788526 ],
[ 968741.83954407],
[ 463714.83552309],
[ 458749.5202625 ],
[ 467566.06735236],
[ 251228.59990596],
[ 367839.0507926 ],
[ 178563.73689914],
[ 404748.33480203],
[ 361629.76401877],
[ 335003.99532863],
[ 468590.27220151],
[ 342562.21592532],
[ 371236.86365252],
[ 310876.80170877],
[ 276928.20517629],
[ 273106.88139315],
[ 517720.942006 ],
[ 309671.80782539],
[ 321349.71807989],
[ 346542.76153275],
[ 374434.93476654],
[ 919566.5541124 ],
[ 295586.59732644],
[ 394954.99452145],
[ 259707.15616347],
[ 296950.34369535],
[ 390519.03008681],
[ 338032.92115378],
[ 279645.24385936],
[ 394134.53393142],
[ 279107.13851167],
[ 454032.21228709],
[ 959467.06920008],
[ 683669.88249351],
[ 371046.82704552],
[ 576981.26372572],
[ 294170.56215725],
[ 443250.48438223],
[ 283176.45780734],
[ 434794.43417916],
[ 344911.47246756],
[ 272235.14370441],
[ 824501.58608444],
[ 334396.91111258],
[ 350120.43652201],
[ 253527.41306202],
[ 293854.69267028],
[ 291547.17430427],
[ 270927.11434524],
[ 383959.89024261],
[ 656199.3017448 ],
[ 553341.638875 ],
[ 432687.22009017],
[ 526139.69394388],
[ 447026.81639178],
[ 274087.19936264],
[ 397721.29867274],
[ 491976.34496209],
[ 334112.72222878],
[ 564079.20871881],
[ 262283.77968584],
[ 325888.17562753],
[ 707612.5028916 ],
[ 392111.24832978],
[ 498668.84201377],
[ 287777.33771682],
[ 269508.93047261],
[ 438284.33540911],
[ 325678.41722469],
[ 451638.79768381],
[ 313355.0076554 ],
[ 507395.87489522],
[ 319663.80512625],
[ 334586.40535677],
[ 336983.25994216],
[ 241704.55116646],
[ 551645.5514437 ],
[ 263184.95608194],
[ 353876.49260028],
[ 261995.60845714],
[ 310876.80170877],
[ 350329.83055109],
[ 322549.60455611],
[ 394105.09522405],
[ 553247.67298293],
[ 318793.17233144],
[ 467137.74037172],
[ 242785.12035314],
[ 382352.68146455],
[ 327735.05085456],
[ 517343.1516039 ],
[ 290640.3481382 ],
[ 219340.07698546],
[ 298372.97925164],
[ 194674.57986053],
[ 273246.9321801 ],
[ 332307.2208042 ],
[ 264140.01584633],
[ 457026.61216604],
[ 395940.59788713],
[ 290989.89791922],
[ 473556.78554839],
[ 318678.94228073],
[ 485526.48500272],
[ 325262.02533913],
[ 481691.81794123],
[ 339771.01794269],
[ 367305.38535593],
[ 443173.11536588],
[ 505575.09158582],
[ 318678.94228073],
[ 216558.78229908],
[ 233490.17904287],
[ 394075.65651668],
[ 295760.63169678],
[ 258795.98327875],
[ 346353.26728856],
[ 305592.11589469],
[ 443578.56014568],
[ 300577.82651562],
[ 402095.62514628],
[ 513536.18280069],
[ 757593.69577602],
[ 287777.33771682],
[ 397881.4272336 ],
[ 414604.91098864],
[ 349552.42312819],
[ 534205.58125196],
[ 378854.88856884],
[ 348203.48930532],
[ 333984.13719815],
[ 330351.10957292],
[ 182753.04098048],
[ 329142.59091077],
[ 290053.63304579],
[ 309309.5704895 ],
[ 481568.04896803],
[ 270985.99175998],
[ 327735.05085456],
[ 237495.07611884],
[ 295905.1341674 ],
[ 267893.10128127],
[ 305518.51912627],
[ 289640.4947576 ],
[ 448016.13092096],
[ 448276.23951762],
[ 306562.53056792],
[ 763649.68167691],
[ 302348.80199401],
[ 316359.12444587],
[ 290899.54999829],
[ 231743.82621443],
[ 313404.81548641],
[ 278816.39312135],
[ 383639.70614492],
[ 300989.76671752],
[ 320213.59809987],
[ 327790.31029817],
[ 365619.94174106],
[ 405226.27145632],
[ 351115.20266379],
[ 301726.6340505 ],
[ 239057.85565444],
[ 623907.54383236],
[ 325824.388963 ],
[ 305965.91217926],
[ 349984.19009093],
[ 638635.09324109],
[ 624904.92801633],
[ 421648.73037838],
[ 290610.1806833 ],
[ 288601.57055477],
[ 504376.11184616],
[ 246333.26344267],
[ 390474.87202575],
[ 676486.10550918],
[ 217145.59058386],
[ 253652.38012152],
[ 874149.26336545],
[ 241282.41631858],
[ 300762.91295673],
[ 364474.75217731],
[ 819789.82291398],
[ 227915.8527324 ],
[ 333171.00574275],
[ 406022.1093964 ],
[ 424620.79041928],
[ 433925.27065205],
[ 369693.34834663],
[ 659209.432679 ],
[ 377446.15042634],
[ 489887.75954764],
[ 215155.31778014],
[ 369953.55013564],
[ 541589.4962969 ],
[ 270941.83369892],
[ 342407.81015845],
[ 259234.77608678],
[ 433939.99000574],
[ 277310.70667081],
[ 675738.22998601],
[ 376222.07869798],
[ 381358.95913228],
[ 366112.32656763],
[ 311543.39889473],
[ 236978.24663136],
[ 321370.9939401 ],
[ 412013.97636694],
[ 321355.16969248],
[ 318664.22292705],
[ 382338.23329227],
[ 581803.00342058],
[ 513506.74409331],
[ 635007.71538581],
[ 270912.39499155],
[ 316897.22979356],
[ 352344.72600475],
[ 256796.27651754],
[ 453674.42663487],
[ 321897.63353148],
[ 307478.51951008],
[ 268956.93948377],
[ 557884.64129867],
[1047884.98951607],
[ 440857.06977894],
[ 226951.61841929],
[ 350348.53224967],
[ 443134.68009882],
[ 277414.30467774],
[ 476949.14679571],
[ 286093.7277434 ],
[ 585186.00373445],
[ 441842.20380586],
[ 265771.94046669],
[ 318649.50357336],
[ 252423.12796196],
[ 321470.14026336],
[ 316360.32253217],
[ 318678.94228073],
[ 249636.38166299],
[ 342591.65463269],
[ 255418.53954247],
[ 911134.08274977],
[ 427920.19747611],
[ 528632.81740156],
[1170092.55572165],
[ 394119.81457774],
[ 306503.65315318],
[ 312892.06153619],
[ 233843.14863766],
[ 490663.22836409],
[ 459535.36171397],
[ 333969.41784446],
[ 282679.25692334],
[ 226111.82057552],
[ 538215.56556676],
[ 226111.82057552],
[ 352457.85116153],
[-161201.02176254],
[ 302303.81022042],
[ 323007.37024413],
[ 370455.47388471],
[ 287524.71253158],
[1060592.82741372],
[ 248909.9801574 ],
[ 288108.35201566],
[ 287812.79056791],
[ 229678.75855599],
[ 249386.81191776],
[ 390314.04386082],
[ 451537.07719892],
[ 528632.81740156],
[ 811858.17040649],
[ 302786.19855838],
[ 394915.92369922],
[ 390314.04386082],
[ 405425.39781537],
[1289225.09943457],
[ 349637.21447153],
[ 309215.24022367],
[ 347266.58887672],
[ 295166.06881919],
[ 301125.47078581],
[ 738397.19827818],
[ 494697.19977366],
[ 305130.09668038],
[ 278844.04749905],
[ 302727.32114364],
[ 310444.76356463],
[ 430134.21928882],
[ 309671.80782539],
[ 721669.74395088],
[ 597337.30576723],
[ 351846.05585305],
[ 704587.65260372],
[ 584703.8865779 ],
[ 293576.41296521],
[ 252313.71396867],
[ 718524.74266092],
[ 253189.43400231],
[ 302168.8466723 ],
[ 288735.60719799],
[ 498483.16389807],
[ 457172.49071198],
[ 333178.1247804 ],
[ 704558.21389635],
[ 468988.05558085],
[ 417037.85398029],
[ 245048.7518395 ],
[ 423113.73966605],
[ 354931.43920809],
[ 720067.89359305],
[ 749936.43382107],
[ 551719.98192466],
[ 441294.46634331],
[ 556690.74879784],
[ 592416.24176006],
[ 422202.56678133],
[ 318693.39045302],
[ 403430.30895423],
[ 604406.85270084],
[ 255244.59836449],
[ 374597.01387348],
[ 308150.766466 ],
[ 604268.00000018],
[ 616438.10869654],
[ 993142.48503895],
[ 395901.52706491],
[ 434344.3298916 ],
[ 485469.25484471],
[ 627355.43570469],
[ 310876.80170877],
[ 174618.27976904],
[ 447061.07115658],
[ 335585.89436361],
[ 327421.4232634 ],
[ 373803.94825239],
[ 373789.2288987 ],
[ 348443.52012806],
[ 372500.73495064],
[ 556630.4021154 ],
[ 553112.80262718],
[ 400867.67603802],
[ 569468.03490261],
[ 444493.62218294],
[ 514644.81461941],
[ 702403.53883714],
[ 608248.81678902],
[1162131.46450679],
[ 314668.1242534 ],
[ 344572.38851392],
[ 641419.64152483],
[ 406890.71039238],
[ 330256.77930708],
[ 406762.86588192],
[ 453663.96080749],
[1656917.6302292 ],
[ 449408.85843111],
[ 258707.30278287],
[ 345745.18314357],
[ 310109.76692089],
[ 424505.72665604],
[ 310095.0475672 ],
[ 401125.77300418],
[ 457988.13524458],
[ 550575.62607344],
[ 523303.70234502],
[ 506600.1301475 ],
[ 379014.11039314],
[ 352767.4964078 ],
[ 644515.6686963 ],
[ 537987.4698391 ]])y_test = pd.read_csv("./dev-0/expected.tsv", header=None)
y_test| 0 | |
|---|---|
| 0 | 373000.00 |
| 1 | 299000.00 |
| 2 | 365000.00 |
| 3 | 369000.00 |
| 4 | 483791.00 |
| ... | ... |
| 457 | 655544.02 |
| 458 | 471397.97 |
| 459 | 309958.00 |
| 460 | 699000.00 |
| 461 | 850000.00 |
462 rows × 1 columns
r2_score(y_test, results)0.6393762535622007
mean_absolute_error(y_test, results)71559.96181964973
Predykcja dla zbioru testowego
final_test_dataset = pd.read_csv("./test-A/in.tsv", sep= "\t", header=None)
# Preprocessing column 15:
print("COLUMN 15:")
# Count the occurrence of unique values in column before preprocessing:
print("Value counts before changes:\n",final_test_dataset[14].value_counts())
# Replace string to int or NaN:
final_test_dataset[14] = final_test_dataset[14].replace({"parter": 0, "suterena": -1, "> 10": 11, "poddasze": np.nan})
final_test_dataset[14] = final_test_dataset[14].apply(float)
# Fill Nans with median:
final_test_dataset[14].fillna(final_test_dataset[14].median(), inplace=True)
final_test_dataset[14]= final_test_dataset[14].apply(int)
# Count the occurrence of unique values in column after preprocessing:
print("Value counts after changes:\n",final_test_dataset[14].value_counts())
# Preprocessing column 8:
print("COLUMN 8:")
# Replace strings containing space to NaN:
final_test_dataset[7] = final_test_dataset[7].replace(' ', np.nan, regex=True)
# Fill Nans with median:
final_test_dataset[7] = final_test_dataset[7].apply(float)
final_test_dataset[7].fillna(final_test_dataset[7].median(), inplace=True)
print(final_test_dataset[7])
# Preprocessing column 6:
print("COLUMN 6:")
# Count the occurrence of unique values in column before preprocessing:
print("Value counts before changes:\n",final_test_dataset[5].value_counts())
# Change string to 10:
final_test_dataset[5] = final_test_dataset[5].replace({"więcej niż 10": 10})
final_test_dataset[5] = final_test_dataset[5].apply(int)
# Count the occurrence of unique values in column after preprocessing:
print("Value counts after changes:\n",final_test_dataset[5].value_counts())
final_test_dataset[9].fillna(final_test_dataset[9].median(), inplace=True)
final_test_dataset[9] = final_test_dataset[9].apply(float)
final_test_dataset = final_test_dataset[[5,7,9,14]]
COLUMN 15:
Value counts before changes:
14
1 92
parter 70
3 68
4 64
2 61
5 15
6 11
7 7
10 5
> 10 5
9 4
8 2
suterena 1
Name: count, dtype: int64
Value counts after changes:
14
1 92
2 74
0 70
3 68
4 64
5 15
6 11
7 7
10 5
11 5
9 4
8 2
-1 1
Name: count, dtype: int64
COLUMN 8:
0 61.99
1 64.00
2 51.15
3 45.77
4 44.36
...
413 34.97
414 49.06
415 76.71
416 72.63
417 65.84
Name: 7, Length: 418, dtype: float64
COLUMN 6:
Value counts before changes:
5
2 175
3 143
4 50
1 40
5 6
6 2
więcej niż 10 1
8 1
Name: count, dtype: int64
Value counts after changes:
5
2 175
3 143
4 50
1 40
5 6
6 2
10 1
8 1
Name: count, dtype: int64
final_test_dataset| 5 | 7 | 9 | 14 | |
|---|---|---|---|---|
| 0 | 3 | 61.99 | 7.0 | 2 |
| 1 | 4 | 64.00 | 4.0 | 0 |
| 2 | 3 | 51.15 | 5.0 | 0 |
| 3 | 2 | 45.77 | 7.0 | 2 |
| 4 | 2 | 44.36 | 13.0 | 5 |
| ... | ... | ... | ... | ... |
| 413 | 1 | 34.97 | 8.0 | 4 |
| 414 | 3 | 49.06 | 3.0 | 3 |
| 415 | 3 | 76.71 | 5.0 | 3 |
| 416 | 3 | 72.63 | 5.0 | 3 |
| 417 | 2 | 65.84 | 10.0 | 3 |
418 rows × 4 columns
final_results = reg.predict(final_test_dataset)pd.DataFrame(final_results)| 0 | |
|---|---|
| 0 | 426282.351904 |
| 1 | 389890.897311 |
| 2 | 334372.288463 |
| 3 | 341143.667679 |
| 4 | 346709.875023 |
| ... | ... |
| 413 | 301974.734528 |
| 414 | 312195.369919 |
| 415 | 537901.937976 |
| 416 | 505420.685819 |
| 417 | 509311.081663 |
418 rows × 1 columns
pd.DataFrame(final_results).to_csv("./test-A/out.tsv", sep='\t', index=False, header=None)