
# Loading libraries

from scipy.stats import chi2_contingency
import os
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import warnings
warnings.filterwarnings('ignore')


from google.colab import drive 
drive.mount('/content/drive')

Mounted at /content/drive


train_df = pd.read_csv('/content/drive/MyDrive/Colab Notebooks/Kaggle prac/TPS-mar-2021/train.csv')
test_df = pd.read_csv('/content/drive/MyDrive/Colab Notebooks/Kaggle prac/TPS-mar-2021/test.csv')


sample_submission = pd.read_csv('/content/drive/MyDrive/Colab Notebooks/Kaggle prac/TPS-mar-2021/sample_submission.csv')


train_df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 300000 entries, 0 to 299999
Data columns (total 32 columns):
 #   Column  Non-Null Count   Dtype  
---  ------  --------------   -----  
 0   id      300000 non-null  int64  
 1   cat0    300000 non-null  object 
 2   cat1    300000 non-null  object 
 3   cat2    300000 non-null  object 
 4   cat3    300000 non-null  object 
 5   cat4    300000 non-null  object 
 6   cat5    300000 non-null  object 
 7   cat6    300000 non-null  object 
 8   cat7    300000 non-null  object 
 9   cat8    300000 non-null  object 
 10  cat9    300000 non-null  object 
 11  cat10   300000 non-null  object 
 12  cat11   300000 non-null  object 
 13  cat12   300000 non-null  object 
 14  cat13   300000 non-null  object 
 15  cat14   300000 non-null  object 
 16  cat15   300000 non-null  object 
 17  cat16   300000 non-null  object 
 18  cat17   300000 non-null  object 
 19  cat18   300000 non-null  object 
 20  cont0   300000 non-null  float64
 21  cont1   300000 non-null  float64
 22  cont2   300000 non-null  float64
 23  cont3   300000 non-null  float64
 24  cont4   300000 non-null  float64
 25  cont5   300000 non-null  float64
 26  cont6   300000 non-null  float64
 27  cont7   300000 non-null  float64
 28  cont8   300000 non-null  float64
 29  cont9   300000 non-null  float64
 30  cont10  300000 non-null  float64
 31  target  300000 non-null  int64  
dtypes: float64(11), int64(2), object(19)
memory usage: 73.2+ MB


df = pd.concat([train_df, test_df], sort=False)


df.describe()


label = df['id']

df = df.drop(['id'], axis = 1)


df.isna().sum() + df.isnull().sum()

cat0           0
cat1           0
cat2           0
cat3           0
cat4           0
cat5           0
cat6           0
cat7           0
cat8           0
cat9           0
cat10          0
cat11          0
cat12          0
cat13          0
cat14          0
cat15          0
cat16          0
cat17          0
cat18          0
cont0          0
cont1          0
cont2          0
cont3          0
cont4          0
cont5          0
cont6          0
cont7          0
cont8          0
cont9          0
cont10         0
target    400000
dtype: int64


num_columns = df.drop('target', axis=1).select_dtypes(exclude = ['object']).columns
categ_columns = df.select_dtypes(include = ['object']).columns


num_columns

Index(['cont0', 'cont1', 'cont2', 'cont3', 'cont4', 'cont5', 'cont6', 'cont7',
       'cont8', 'cont9', 'cont10'],
      dtype='object')


categ_columns

Index(['cat0', 'cat1', 'cat2', 'cat3', 'cat4', 'cat5', 'cat6', 'cat7', 'cat8',
       'cat9', 'cat10', 'cat11', 'cat12', 'cat13', 'cat14', 'cat15', 'cat16',
       'cat17', 'cat18'],
      dtype='object')


from sklearn.preprocessing import LabelEncoder


le = LabelEncoder()
for col in categ_columns.tolist():
  df[col] = le.fit_transform(df[col])


df.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 500000 entries, 0 to 199999
Data columns (total 31 columns):
 #   Column  Non-Null Count   Dtype  
---  ------  --------------   -----  
 0   cat0    500000 non-null  int64  
 1   cat1    500000 non-null  int64  
 2   cat2    500000 non-null  int64  
 3   cat3    500000 non-null  int64  
 4   cat4    500000 non-null  int64  
 5   cat5    500000 non-null  int64  
 6   cat6    500000 non-null  int64  
 7   cat7    500000 non-null  int64  
 8   cat8    500000 non-null  int64  
 9   cat9    500000 non-null  int64  
 10  cat10   500000 non-null  int64  
 11  cat11   500000 non-null  int64  
 12  cat12   500000 non-null  int64  
 13  cat13   500000 non-null  int64  
 14  cat14   500000 non-null  int64  
 15  cat15   500000 non-null  int64  
 16  cat16   500000 non-null  int64  
 17  cat17   500000 non-null  int64  
 18  cat18   500000 non-null  int64  
 19  cont0   500000 non-null  float64
 20  cont1   500000 non-null  float64
 21  cont2   500000 non-null  float64
 22  cont3   500000 non-null  float64
 23  cont4   500000 non-null  float64
 24  cont5   500000 non-null  float64
 25  cont6   500000 non-null  float64
 26  cont7   500000 non-null  float64
 27  cont8   500000 non-null  float64
 28  cont9   500000 non-null  float64
 29  cont10  500000 non-null  float64
 30  target  300000 non-null  float64
dtypes: float64(12), int64(19)
memory usage: 122.1 MB


train_df = df.iloc[:len(train_df)]
test_df = df.iloc[len(train_df):]

X = train_df.drop('target', axis=1)
y = train_df['target']

y

0         0.0
1         0.0
2         0.0
3         0.0
4         1.0
         ... 
299995    0.0
299996    0.0
299997    1.0
299998    0.0
299999    0.0
Name: target, Length: 300000, dtype: float64


y.value_counts()

0.0    220539
1.0     79461
Name: target, dtype: int64


from sklearn.model_selection import train_test_split

X_train, X_valid, y_train, y_valid = train_test_split(X, y, test_size=0.2, random_state=42, stratify=y)


params = {
    "eta": 0.1,
    "objective": "binary:logistic", 
    "eval_metric": "logloss",
    "learning_rate" : 0.05,
    "max_depth": 4,
    "min_child_weight": 2
}
num_rounds = 100


# learning rate : 0.05 (0.8836) -> 0.1 (inc) -> 0.15 (inc) -> 0.2 (inc)
# num_iter : 100 -> 500 (inc) -> 1000(dec) -> 600 (no change) -> 400 (inc) -> 300 (dec)
import lightgbm as lgb

lgb_model = lgb.LGBMClassifier(eta = 0.1, eval_metric = "logloss", learning_rate = 0.2,
                               max_depth = 4, min_child_weight = 1, metric = "auc", num_iterations = 400)
lgb_model.fit(X_train, y_train)

LGBMClassifier(eta=0.1, eval_metric='logloss', learning_rate=0.2, max_depth=4,
               metric='auc', min_child_weight=1, num_iterations=400)


y_pred = lgb_model.predict(X_valid)
y_pred_prob = lgb_model.predict_proba(X_valid)


pred = pd.DataFrame({'target':y_valid, 'pred':y_pred})
pred_proba = pd.DataFrame({'target':y_valid, 'class0_proba':y_pred_prob[:, 0], 'class1_proba':y_pred_prob[:, 1]})


from sklearn.metrics import accuracy_score, precision_score, recall_score
from sklearn.metrics import f1_score, roc_auc_score

accuracy = accuracy_score(y_valid, y_pred)
precision = precision_score(y_valid, y_pred)
recall = recall_score(y_valid, y_pred)
f1 = f1_score(y_valid, y_pred)
auc = roc_auc_score(y_valid, y_pred_prob[:, 1])
   
print(f"정확도: {accuracy:.4f}, 정밀도: {precision:.4f}, 재현율: {recall:.4f}, F1: {f1:.4f}, AUC: {auc:.4f}")

정확도: 0.8488, 정밀도: 0.7629, 재현율: 0.6228, F1: 0.6858, AUC: 0.8906


s_pred = lgb_model.predict(test_df.drop('target', axis=1))
s_pred_prob = lgb_model.predict_proba(test_df.drop('target', axis=1))[:,1]


print(len(sample_submission), )
print(len(s_pred), s_pred)
print(len(s_pred_prob), s_pred_prob)

200000
200000 [0. 0. 0. ... 1. 0. 0.]
200000 [0.09709812 0.39236477 0.03227922 ... 0.64365668 0.15505113 0.45039296]


submission = s_pred_prob
id_col = label.iloc[len(train_df):]


submission_df = pd.DataFrame(id_col)
submission_df['target'] = submission
submission_df.head()


submission_df.to_csv('submission.csv', index=False)

	id	cont0	cont1	cont2	cont3	cont4	cont5	cont6	cont7	cont8	cont9	cont10	target
count	500000.000000	500000.000000	500000.000000	500000.000000	500000.000000	500000.000000	500000.000000	500000.000000	500000.000000	500000.000000	500000.000000	500000.000000	300000.000000
mean	249999.500000	0.505004	0.494028	0.516549	0.474655	0.504817	0.501827	0.487863	0.502117	0.488177	0.469623	0.508528	0.264870
std	144337.711635	0.207040	0.212900	0.214811	0.216750	0.227364	0.241296	0.211335	0.203615	0.178987	0.194709	0.203606	0.441265
min	0.000000	-0.049562	0.084480	0.094493	-0.045316	0.168071	-0.036379	-0.005648	0.090901	0.013554	0.214866	0.097789	0.000000
25%	124999.750000	0.344881	0.317528	0.326111	0.292455	0.279346	0.276002	0.323335	0.353192	0.359006	0.309916	0.368589	0.000000
50%	249999.500000	0.479646	0.463183	0.471095	0.389550	0.480037	0.554752	0.517779	0.435724	0.411067	0.407304	0.446577	0.000000
75%	374999.250000	0.638083	0.693951	0.704678	0.645278	0.725551	0.735190	0.608117	0.643630	0.612634	0.587243	0.582422	1.000000
max	499999.000000	1.006825	1.011597	1.016600	0.965609	0.867198	0.853022	0.966553	1.035818	1.055885	1.007189	1.011331	1.000000

TPS-mar-2021 Report

TPS-mar-2021 Data

Purpose & Contents

Data

Summary

TPS-mar-2021 Report

TPS-mar-2021 Report

TPS-mar-2021 Data

Purpose & Contents

Data

Summary

Code

TAEWON KIM

Recent posts

Comments

	id	target
0	5	0.097098
1	6	0.392365
2	8	0.032279
3	9	0.208910
4	11	0.202885

TPS-mar-2021 Report

TPS-mar-2021 Data

Purpose & Contents

Data

Summary

TPS-mar-2021 Report

TPS-mar-2021 Report

TPS-mar-2021 Data

Purpose & Contents

Data

Summary

Code

TAEWON KIM

Recent posts

Newsletter

Comments