使用apply python pandas生成多个列

Question

我有一个返回字典的函数，我想使用apply将新列添加到现有数据框

import pandas as pd
from chainmap import ChainMap

# features1 is a dictionary returned from feature_func1
#features2 is a dictionary returned from feature_func2
def applyfeatures(x,datacolumn1,datacolumn2):
    features1=feature_func1(x[datacolumn1],x[datacolumn2])
    features2=feature_func2(x[datacolumn1],x[datacolumn2])
    z= ChainMap(features1,features2)
    return z

if __name__ == '__main__':
    df=pd.read_csv(input_file)
    datacolumn1="data1"
    datacolumn2="data2"
    newdf=df.apply(applyfeatures,args=(datacolumn1,datacolumn2),axis=1)
    merged=df.join(newdf)

这会出错“其他系列必须有名称”。这可能是由于我从applyfeatures返回字典的方式。我应该如何修改它以基于返回的字典项创建新的数据帧。