Commit Live PR

q01_missing_value/build.py

@@ -1,10 +1,25 @@
+# %load q01_missing_value/build.py
 # Default imports
 import pandas as pd
-
+from sklearn.preprocessing import Imputer
 # Data loading
 ny_housing = pd.read_csv('data/train.csv')
 # Selecting 4 most relevant variables along with target variable from the dataset fot the Cleaning and Preprocessing.
 housing_data = ny_housing[['MasVnrArea', 'GrLivArea', 'LotShape', 'GarageType', 'SalePrice']]
 
+def imputation(df):
+    numeric_features = [a for a in range(len(df.dtypes)) if df.dtypes[a] in ['int64','float64']]
+    numeric_df = df.iloc[:, numeric_features]
+    cat_features = df.columns.difference(df.columns[numeric_features])
+    cat_df = df.loc[:,cat_features]
+    numeric_imputer = Imputer(missing_values = 'NaN', strategy='mean')
+    numeric_imputed_df = pd.DataFrame(numeric_imputer.fit_transform(numeric_df))
+    numeric_imputed_df.columns = numeric_df.columns
+    numeric_imputed_df.index = numeric_df.index
+    for feature in cat_features:
+        cat_df[feature] = cat_df[feature].fillna(cat_df[feature].mode()[0])
+    return numeric_imputed_df, cat_df
+
+
+
 
-# Write your code here:

q02_outlier_removal/build.py

@@ -1,3 +1,4 @@
+# %load q02_outlier_removal/build.py
 # Default imports
 import pandas as pd
 
@@ -6,5 +7,16 @@
 # Selecting 4 most relevant variables from the dataset fot the Cleaning and Preprocessing.
 housing_data = ny_housing[['MasVnrArea', 'GrLivArea', 'LotShape', 'GarageType', 'SalePrice']]
 
+def outlier_removal(df):
+    qv = 0.95
+    df_qv = df.quantile(q=qv, axis=0, numeric_only=True, interpolation='linear')
+    numeric_features = [a for a in range(len(df.dtypes)) if df.dtypes[a] in ['int64','float64']]
+    numeric_df = df.iloc[:, numeric_features]
+    for feature in numeric_df.columns:
+        df=df.drop(df[df[feature]>df_qv[feature]].index)
+    return df
+
+
+
+
 
-# Write your code here:

q03_skewness_log/build.py

@@ -1,8 +1,15 @@
+# %load q03_skewness_log/build.py
 from scipy.stats import skew
 import pandas as pd
 import numpy as np
 
 data = pd.read_csv('data/train.csv')
 
+def skewness_log(df):
+    df_trans = df.copy()
+    df_trans['SalePrice'] = np.log(df_trans['SalePrice'])
+    df_trans['GrLivArea'] = np.log(df_trans['GrLivArea'])
+    return skew(df_trans['GrLivArea']), skew(df_trans['SalePrice'])
+
+
 
-# Write code here:

q03_skewness_sqrt/build.py

@@ -1,10 +1,16 @@
+# %load q03_skewness_sqrt/build.py
 # Default imports
 from scipy.stats import skew
 import pandas as pd
 import numpy as np
 
 ny_housing = pd.read_csv('data/train.csv')
 
+def skewness_sqrt(df):
+    df_trans = df.copy()
+    df_trans['SalePrice'] = np.sqrt(df_trans['SalePrice'])
+    df_trans['GrLivArea'] = np.sqrt(df_trans['GrLivArea'])
+    return skew(df_trans['GrLivArea']), skew(df_trans['SalePrice'])
+
 
-# Write your Solution Here:
 

q04_encoding/build.py

@@ -1,10 +1,17 @@
+# %load q04_encoding/build.py
 # Default imports
 import pandas as pd
 from sklearn.preprocessing import LabelEncoder
 
 ny_housing = pd.read_csv('data/train.csv')
 housing_data = ny_housing[['MasVnrArea', 'GrLivArea', 'LotShape', 'GarageType', 'SalePrice']]
 
+def encoding(df):
+    le = LabelEncoder()
+    df['LotShape_Label'] = le.fit_transform(df['LotShape'])
+    df=df.join(pd.get_dummies(df['GarageType'], drop_first=True))
+    return df
+
+
 
-# Write your code here: