인프런 커뮤니티 질문&답변

#1. 문제파악
# 분류모델 / ROC-AUC, 정확도(Accuracy), F1, 정밀도(Precision), 재현율(Recall)
#TARGET : Attrition_Flag
# CSV("CLIENTNUM,Attrition_Flag) , 수험번호.CSV


#2. 데이터불러오기
import pandas as pd

train = pd.read_csv("https://raw.githubusercontent.com/lovedlim/inf/main/p2/ceredit%20card/train.csv")
test =  pd.read_csv("https://raw.githubusercontent.com/lovedlim/inf/main/p2/ceredit%20card/test.csv")
# train = pd.read_csv("train.csv")
# test =  pd.read_csv("test.csv")
#없을 경우 -> 원핫인코딩 / 있으면 레이블인코딩

#3. EDA
#print(train.shape, test.shape)
#print(train.info(), test.info())
#print(train.isnull().sum(), test.isnull().sum())
#print("test에만 있는 컬럼:", test_cols - train_cols)
#train_cols = set(train.columns)
#test_cols = set(test.columns)



#4. 데이터 전처리
#공통
print(train.shape, test.shape)
target = train.pop('Attrition_Flag')
train =train.drop('CLIENTNUM', axis=1)
id = test.pop('CLIENTNUM')
print(train.shape, test.shape)


#레이블 인코딩
cols = train.select_dtypes(include = 'object').columns
from sklearn.preprocessing import LabelEncoder
for col in cols  :
     le = LabelEncoder()
     train[col] = le.fit_transform(train[col])

     test[col]= le.transform(test[col])


#5. 데이터 분할
from sklearn.model_selection import train_test_split
X_tr, X_val, y_tr, y_val = train_test_split(train, target, test_size=0.2, random_state=2)

#6. 모델 학습 및 평가
from sklearn.ensemble import RandomForestClassifier
rf = RandomForestClassifier(random_state=2)
rf.fit(X_tr,y_tr)
pred = rf.predict(X_val)
pred1 = rf.predict_proba(X_val)

from sklearn.metrics import roc_auc_score, f1_score, recall_score, accuracy_score, precision_score
roc = roc_auc_score(y_val, pred1[:,1])
f1 = f1_score(y_val, pred)
recall = recall_score(y_val, pred)
acc = accuracy_score(y_val, pred)
pre = precision_score(y_val, pred)




#7 평가
pred_test = rf.predict(test)
pred1 = rf.predict_proba(test)
#print(pred1)
submit = pd.DataFrame({'CLIENTNUM' : id, 'Attrition_Flag' : pred1[:,1]})
submit.to_csv("0000.csv", index = False)

pd.read_csv("0000.csv")

안녕하세요~
이번엔 레이블인코딩으로 데이터전처리해서 풀어봤습니다 혹시 피드백해주실 부분이 있으실까요~?