Commit Live PR

q01_pipeline/build.py

@@ -3,11 +3,51 @@
 from sklearn.model_selection import train_test_split, GridSearchCV
 from sklearn.preprocessing import LabelEncoder
 from sklearn.ensemble import RandomForestClassifier
+from sklearn.ensemble import GradientBoostingClassifier
+from sklearn.feature_selection import RFE
+from imblearn.over_sampling import SMOTE
+from sklearn.linear_model import LogisticRegression
 from sklearn.utils.class_weight import compute_class_weight
 from sklearn.metrics import roc_auc_score
 
 bank = pd.read_csv('data/Bank_data_to_class.csv', sep=',')
 
+y = bank['y']
+X = bank.drop('y', axis=1)
+X_train,X_test,y_train,y_test = train_test_split(X,y,test_size=0.3,random_state=9)
+model = GradientBoostingClassifier(random_state=9)
 # Write your solution here :
+def pipeline(X_train,X_test,y_train,y_test,model):
+    bank = pd.read_csv('data/Bank_data_to_class.csv', sep=',')
+    y = bank['y']
+    X = bank.drop('y', axis=1)
+    param_grid = {'n_estimators':[10,20,30], 'max_features':[2,4,6,8,12,16], 'max_depth':[2,4,6,8]}
+    numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']
+    num_features = X_train.select_dtypes(include=numerics).columns.tolist()
+    cat_features = list(set(X_train.columns.tolist()) - set(num_features))
+    onehot_en = ['marital','contact','month','job']
+    label_en = list(set(cat_features) - set(onehot_en))
+    for val in cat_features:
+        temp = X_train[X_train[val] != 'unknown'][val].mode()[0]
+        X_train.loc[X_train[val] == 'unknown',val] = temp
+        X_test.loc[X_test[val] == 'unknown',val] = temp
 
+    encoded_df_train = X_train
+    encoded_df_test = X_test
+    le = LabelEncoder()
+    for val in cat_features:
+        encoded_df_train[val] = le.fit_transform(encoded_df_train[val])
+        encoded_df_test[val] = le.fit_transform(encoded_df_test[val])
+    y_train1 = le.fit_transform(y_train)
+    y_test1 = le.fit_transform(y_test)
+    sm = SMOTE(random_state=9)
+    X_res_train, y_res_train = sm.fit_sample(encoded_df_train,y_train1)
+    X_res_test, y_res_test = sm.fit_sample(encoded_df_test,y_test1)
+    #model.fit(X_res_train,y_res_train)
+    search = GridSearchCV(model, param_grid)
+    search.fit(X_res_train,y_res_train)
+    y_pred = search.predict_proba(X_res_test)[:, 1]
+    auc = roc_auc_score(y_res_test,y_pred)
+    return search, auc
 
+#print(pipeline(X_train,X_test,y_train,y_test,model))