信用評分卡常用函數(二)

• 自定義分箱

df=pd.DataFrame({
    "age":[1.1,2.1,1.3,3.4,5.6,7.8,5.2,3.4,7.4,6.1,6.9,8.9],
    "tag":['good','good','good','bad','good','bad','bad','bad','good','bad','bad','bad']
})


def bin_single_column_selfdef(df,colname,bins,is_distribute=False,labels=None):
    cut_res=pd.cut(df[colname], bins=bins,labels=labels)
    if is_distribute:
        return pd.value_counts(cut_res),cut_res
    else:
        return cut_res

'''
bins=[0,3,6,8,10]
labels=['a','b','c','d']
bin_dis,cut_bin=bin_single_column_selfdef(df,'age',bins,labels,True)
print(bin_dis)
print(cut_bin)
'''

• 等寬

def bin_single_column_byrange(df,colname,bin_len,is_distribute=False,labels=None):
    cut_res=pd.cut(df[colname], bins=bin_len,labels=labels)
    if is_distribute:
        return pd.value_counts(cut_res),cut_res
    else:
        return cut_res

'''
x,y=bin_single_column_byrange(df,'age',3,True)  # 這裏的3表示區間段
print(x)
print(y)
'''

• 等頻

• def bin_single_column_byfreq(df,colname,bin_len,is_distribute=False,labels=None):
      cut_res=pd.qcut(df[colname], q=bin_len,labels=labels)
      if is_distribute:
          return pd.value_counts(cut_res),cut_res
      else:
          return cut_res
  
  '''
  x,y=bin_single_column_byfreq(df,'age',3,True)  # 這裏的3表示分位點，等同於[0,1/3分位點，2/3分位點,1分位點]，所以不一定能保證每個組的個數是一樣的
  print(x)
  print(y)
  '''