aniruddha27

print('Skewness = ',train['SalePrice'].skew())

aniruddha27

target = np.log(train['SalePrice'])
print('Skewness = ',target.skew())
sns.distplot(target);

aniruddha27

corr = train.corr()
corr['SalePrice'].sort_values(ascending=False).head(10)

aniruddha27

table = pd.pivot_table(train,index='OverallQual',values='SalePrice',aggfunc=np.mean)
table

aniruddha27

plt.scatter(x=train['GrLivArea'], y=train['SalePrice'])
plt.ylabel('Sale Price')
plt.xlabel('GrLivArea')
plt.show();

aniruddha27

# dropping outlier values from the dataset
train = train[train['GrLivArea']<4500]

aniruddha27

#train rows
ntrain = train.shape[0]

#save log transform of target feature
target = np.log(train['SalePrice'])

#drop Id and SalePrice from train dataframe
train.drop(['Id','SalePrice'],inplace=True,axis=1)

#store test Id

aniruddha27

#Null values
train.isna().sum().sort_values(ascending=False).head(20)

aniruddha27

#                                                Ordinal features
#NA means no Pool
train['PoolQC'].replace(['Ex','Gd','TA','Fa',np.nan],[4,3,2,1,0],inplace=True)

# NA means no fence
train['Fence'].replace(['GdPrv','MnPrv','GdWo','MnWw',np.nan],[4,3,2,1,0],inplace=True)

# NA means no fireplace
train['FireplaceQu'].replace(['Ex','Gd','TA','Fa','Po',np.nan],[5,4,3,2,1,0],inplace=True)

aniruddha27

#                                                Ordinal features
for i in ['GarageCond','GarageQual']:
    train[i].replace(['Ex','Gd','TA','Fa','Po',np.nan],[5,4,3,2,1,0],inplace=True)
    
#                                                 Nominal features
for i in ['GarageFinish','GarageType']:
    train[i].fillna('None',inplace=True) 
    
#                                               Numerical features
for i in ['GarageYrBlt','GarageCars','GarageArea']: