pipeline
Pipeline in Pandas allows to build a sequence of function to run in order on data-frame.
source ="https://storage.googleapis.com/mledu-datasets/california_housing_train.csv"
CHT = pd.read_csv(source, sep=",")
def categ(x,col):
x[col].quantile(.3)
x['lev'] = ''
C1=x[col]<=x[col].quantile(.3)
C2=x[col]>=x[col].quantile(.7)
x.loc[C1,'famlev']=0
x.loc[~C1&~C2,'famlev']=1
x.loc[C2,'famlev']=2
return x
def cv(x):
return (np.mean(x)/np.var(x))
CHT.pipe(cv)
CHT.pipe(categ, col='median_income').pipe(cv)
