Create house

competitions/getting-started/house

@@ -0,0 +1,98 @@
+import pandas as pd
+import numpy as np
+from sklearn.linear_model import Lasso,LinearRegression,Ridge,ElasticNet,TheilSenRegressor,HuberRegressor,RANSACRegressor
+from sklearn.svm import SVR
+from sklearn.tree import DecisionTreeRegressor,ExtraTreeRegressor
+from sklearn.ensemble import AdaBoostRegressor,ExtraTreesRegressor,GradientBoostingRegressor,RandomForestRegressor
+from xgboost import XGBRegressor
+from sklearn.model_selection import train_test_split
+from sklearn.metrics import mean_squared_error
+import itertools
+train = pd.read_csv('train.csv')
+train.head()
+test = pd.read_csv('test.csv')
+test.head()
+test.dtypes['Id']
+num_columns = []
+cate_columns = []
+for column in test.columns:
+    if test.dtypes[column]!=np.dtype('object'):
+        num_columns.append(column)
+    else:
+        cate_columns.append(column)
+print(num_columns)
+print(cate_columns)
+for column in num_columns:
+    train[column] = train[column].fillna(train[column].median())
+    test[column] = test[column].fillna(test[column].median())
+
+for column in cate_columns:
+    train[column] = train[column].fillna('NaN')
+    test[column] = test[column].fillna('NaN')
+
+data = pd.concat([train,test],axis=0)
+for column in cate_columns:
+    t = pd.get_dummies(data[column],prefix=column)
+    train = pd.concat([train,t[:len(train):]],axis=1)
+    train.drop(column,axis=1,inplace=True)
+    test = pd.concat([test,t[len(train):]],axis=1)
+    test.drop(column,axis=1,inplace=True)
+
+label = train.pop('SalePrice')
+X_train, X_test, y_train, y_test = train_test_split(train, label, test_size=0.33, random_state=42)
+regs = [
+    ['Lasso',Lasso()],
+    ['LinearRegression',LinearRegression()],
+    ['Ridge',Ridge()],
+    ['ElasticNet',ElasticNet()],
+    ['TheilSenRegressor',TheilSenRegressor()],
+    ['RANSACRegressor',RANSACRegressor()],
+    ['HuberRegressor',HuberRegressor()],
+    ['SVR',SVR(kernel='linear')],
+    ['DecisionTreeRegressor',DecisionTreeRegressor()],
+    ['ExtraTreeRegressor',ExtraTreeRegressor()],
+    ['AdaBoostRegressor',AdaBoostRegressor(n_estimators=150)],
+    ['ExtraTreesRegressor',ExtraTreesRegressor(n_estimators=150)],
+    ['GradientBoostingRegressor',GradientBoostingRegressor(n_estimators=150)],
+    ['RandomForestRegressor',RandomForestRegressor(n_estimators=150)],
+    ['XGBRegressor',XGBRegressor(n_estimators=150)],
+]
+preds = []
+for reg_name,reg in regs:
+    print(reg_name)
+    reg.fit(X_train,y_train)
+    y_pred = reg.predict(X_test)
+    if np.sum(y_pred<0) > 0:
+        print('y_pred have', np.sum(y_pred < 0), 'negative values, we fill it with np.median(y_pred)')
+        y_pred[y_pred < 0] = np.median(y_pred)
+    score = np.sqrt(mean_squared_error(np.log(y_test), np.log(y_pred)))
+    preds.append([reg_name, y_pred])
+
+final_results = []
+for comb_length in range(1,len(regs)+1):
+    print('Model Amount:',comb_length)
+    results = []
+    for comb in itertools.combinations(preds,comb_length):
+        pred_sum = 0
+        model_name = []
+        for reg_name,pred in comb:
+            pred_sum += pred
+            model_name.append(reg_name)
+        pred_sum /= comb_length
+        model_name = '+'.join(model_name)
+        score = np.sqrt(mean_squared_error(np.log(y_test),np.log(pred_sum)))
+        results.append([model_name,score])
+    results = sorted(results,key=lambda x:x[1])
+    for model_name,score in results:
+        print(model_name,score)
+    print()
+    final_results.append(results[0])
+
+final_result = sorted(final_results,key=lambda x:x[1])
+for model_name,score in final_results:
+    print(model_name,score)
+
+[b for b in zip(itertools.count(),[a[0] for a in regs])]
+pred = np.mean(list(map(lambda x:regs[x][1].predict(test),[0,12,14])),axis=0)
+sub = pd.DataFrame({'Id':test['Id'],'SalePrice':pred})
+sub.to_csv('submission_Universe_fillNaN.csv',index=None)