Commit Live PR

q01_load_data/build.py

@@ -1,10 +1,13 @@
+# %load q01_load_data/build.py
 # Default imports
 import pandas as pd
 from sklearn.model_selection import train_test_split
-
-
-path = 'data/house_prices_multivariate.csv'
-
+path='data/house_prices_multivariate.csv'
 
 # Write your solution here
+def load_data(path,test_size=0.33,Random_State=9):
+    df=pd.read_csv(path)
+    X_train, X_test, y_train, y_test = train_test_split(df.drop('SalePrice',axis=1), df.SalePrice, test_size=test_size, random_state=Random_State)
+    return df,X_train,X_test,y_train,y_test
+
 

q02_Max_important_feature/build.py

@@ -1,3 +1,4 @@
+# %load q02_Max_important_feature/build.py
 # Default imports
 from greyatomlib.advanced_linear_regression.q01_load_data.build import load_data
 
@@ -6,3 +7,15 @@
 
 
 # Write your code here
+def Max_important_feature(data_set,target_variable='SalePrice',n=4):
+    x=(data_set.corr()[target_variable])
+    x=x.sort_values(ascending=False)
+    return x.index[1:n+1]
+
+
+
+
+Max_important_feature(data_set,'SalePrice')
+data_set.corr()['SalePrice']
+
+

q03_polynomial/build.py

@@ -1,11 +1,23 @@
+# %load q03_polynomial/build.py
 # Default imports
 from greyatomlib.advanced_linear_regression.q01_load_data.build import load_data
 from sklearn.preprocessing import PolynomialFeatures
 from sklearn.pipeline import make_pipeline
 from sklearn.linear_model import LinearRegression
+from sklearn.pipeline import Pipeline
 
 # We have already loaded the data for you
 data_set, X_train, X_test, y_train, y_test = load_data('data/house_prices_multivariate.csv')
 
 
 # Write your solution here
+def polynomial(power=5,random_state=9):
+    model = Pipeline([('poly', PolynomialFeatures(degree=power)),
+                   ('linear', LinearRegression(fit_intercept=False))])
+    model.fit(X_train,y_train)
+    return model
+import numpy as np
+y=np.array([4, 5, 6, 7]).reshape(1, -1)
+
+
+

q04_ridge/build.py

@@ -1,15 +1,42 @@
+# %load q04_ridge/build.py
 # Default imports
 from sklearn.linear_model import Ridge
 import pandas as pd
 import numpy as np
 from sklearn.metrics import mean_squared_error
 from greyatomlib.advanced_linear_regression.q01_load_data.build import load_data
-np.random.seed(9)
 
 # We have already loaded the data for you
 data_set, X_train, X_test, y_train, y_test = load_data('data/house_prices_multivariate.csv')
 
+np.random.seed(9)
+
 
 # Write your solution here
+def ridge(alpha=0.01):
+    clf=Ridge(alpha=alpha, normalize=True)
+    clf.fit(X_train,y_train)
+    y_pred=clf.predict(X_test)
+    return mean_squared_error(clf.predict(X_train),y_train)**0.5,mean_squared_error(y_pred,y_test)**0.5,clf
+
+
+
+ridge()
+clf=Ridge(alpha=100, normalize=False,random_state=9)
+clf.fit(X_train,y_train)
+y_pred=clf.predict(X_train)
+
+mean_squared_error(y_pred,y_train) **0.5
+from sklearn.linear_model import LinearRegression
+
+clf=LinearRegression()
+clf.fit(X_train,y_train)
+y_pred=clf.predict(X_train)
+
+mean_squared_error(y_pred,y_train) 
+print(y_test.shape,y_pred.shape)
+type(y_pred)
+y_pred[10]
+y_train
 
 

q05_lasso/build.py

@@ -1,14 +1,21 @@
+# %load q05_lasso/build.py
 # Default imports
 from sklearn.linear_model import Lasso
 import pandas as pd
 import numpy as np
 from sklearn.metrics import mean_squared_error
 from greyatomlib.advanced_linear_regression.q01_load_data.build import load_data
-np.random.seed(9)
 
 # We have already loaded the data for you
 data_set, X_train, X_test, y_train, y_test = load_data('data/house_prices_multivariate.csv')
 
+np.random.seed(9)
+
 
 # Write your solution here
+def lasso(alpha=0.01):
+    clf=Lasso(alpha=alpha,normalize=True,random_state=9)
+    clf.fit(X_train,y_train)
+    return mean_squared_error(clf.predict(X_train),y_train)**0.5,mean_squared_error(clf.predict(X_test),y_test)**0.5
+