如何从具有多个列的数据透视表中删除行

由 发布于 pythonpandasdataframe
 收藏

我有一个使用以下代码创建的数据透视表：

pivot_2 = pd.pivot_table(pre_post_df_2, columns = ['pre_post_nota_ind','Party'],  index = ['uni_const_code'], values = ['votesh_percof_total']).reset_index()
pivot_2.rename(columns = {1:'post', 0:'pre'}, inplace = True) 
#I wrote the above code because pre_post_nota_ind takes 0 which represents pre and 1 which represents post

我得到一个数据透视表（数据框类型），看起来像这样：

                    votesh_percof_total
pre_post_nota_ind    pre              post
uni_const_code  IND        IND        NOTA
0|10|1      71.932234   53.922378    33.827091
0|10|10     46.052632   NaN          40.106277
0|10|11     45.171340   28.443649    32.434116
0|10|12     63.429439   30.115420    NaN
0|10|13     30.000000   NaN          46.598414

旁注：第三和第四列均与“后期”期间相关，因此它们具有相同的标题。

Basically, there are sub-columns in the pivot table above. I want to remove those rows for whose value in the NOTA column is Nan

我希望最终输出像：

                    votesh_percof_total
pre_post_nota_ind pre           post
uni_const_code  IND        IND        NOTA
0|10|1      71.932234   53.922378    33.827091
0|10|10     46.052632   NaN          40.106277
0|10|11     45.171340   28.443649    32.434116
0|10|13     30.000000   NaN          46.598414

I tried the following: pivot_2.dropna(subset = pivot_2['NOTA']) and also pivot_2 = pivot_2.votesh_percof_total.post.notnull()]

他们两个都抛出错误。有人可以帮我吗？

评论