python - 连接两个 Pandas 数据框并重新排序列

Question

我有两个数据框(df1 和 df2，如下所示)，它们的列在顺序和计数上都不同。我需要将这两个数据框附加到一个 Excel 文件，其中的列顺序必须按照下面的 Col_list 中指定。

df1 是:

 durable_medical_equipment    pcp  specialist  diagnostic  imaging  generic  formulary_brand  non_preferred_generic  emergency_room  inpatient_facility  medical_deductible_single  medical_deductible_family  maximum_out_of_pocket_limit_single  maximum_out_of_pocket_limit_family plan_name      pdf_name
0                      False  False       False       False    False    False            False                  False           False               False                      False                      False                               False                               False   ABCBCBC  adjnajdn.pdf

... df2 是:

   pcp  specialist  generic  formulary_brand  emergency_room  urgent_care  inpatient_facility  durable_medical_equipment  medical_deductible_single  medical_deductible_family  maximum_out_of_pocket_limit_single  maximum_out_of_pocket_limit_family plan_name      pdf_name
0  True        True    False            False            True         True                True                       True                       True                       True                                True                                True   ABCBCBC  adjnajdn.pdf

我正在创建一个列列表，它与 excel 中列的顺序相同。

Col_list = ['durable_medical_equipment', 'pcp', 'specialist', 'diagnostic',
            'imaging', 'generic', 'formulary_brand', 'non_preferred_generic',
            'emergency_room', 'inpatient_facility', 'medical_deductible_single',
            'medical_deductible_family', 'maximum_out_of_pocket_limit_single', 'maximum_out_of_pocket_limit_family',
            'urgent_care', 'plan_name', 'pdf_name']

我正在尝试使用 concat() 根据 Col_list 重新排序我的数据框。对于数据框中不存在的列值，该值可以是 NaN。

result = pd.concat([df, pd.DataFrame(columns=list(Col_list))])

这不能正常工作。我怎样才能实现这种重新排序？

我尝试了以下方法:

 result = pd.concat([df_repo, pd.DataFrame(columns=list(Col_list))], sort=False, ignore_index=True)
        print(result.to_string())

我得到的输出是:

 durable_medical_equipment    pcp specialist diagnostic imaging generic formulary_brand non_preferred_generic emergency_room inpatient_facility medical_deductible_single medical_deductible_family maximum_out_of_pocket_limit_single maximum_out_of_pocket_limit_family plan_name      pdf_name urgent_care
0                     False  False      False      False   False   False           False                 False          False              False                     False                     False                              False                              False   ABCBCBC  adjnajdn.pdf         NaN
    pcp specialist generic formulary_brand emergency_room urgent_care inpatient_facility durable_medical_equipment medical_deductible_single medical_deductible_family maximum_out_of_pocket_limit_single maximum_out_of_pocket_limit_family plan_name      pdf_name diagnostic imaging non_preferred_generic
0  True       True   False           False           True        True               True                      True                      True                      True                               True                               True   ABCBCBC  adjnajdn.pdf        NaN     NaN                   NaN

Answer 1

如果需要按列表中的值更改顺序，请添加 DataFrame.reindex并传递给 concat:

df = pd.concat([df1.reindex(Col_list, axis=1), 
                df2.reindex(Col_list, axis=1)], sort=False, ignore_index=True)
print (df)
   durable_medical_equipment    pcp  specialist  diagnostic  imaging  generic  \
0                      False  False       False         0.0      0.0    False   
1                       True   True        True         NaN      NaN    False   

   formulary_brand  non_preferred_generic  emergency_room  inpatient_facility  \
0            False                    0.0           False               False   
1            False                    NaN            True                True   

   medical_deductible_single  medical_deductible_family  \
0                      False                      False   
1                       True                       True   

   maximum_out_of_pocket_limit_single  maximum_out_of_pocket_limit_family  \
0                               False                               False   
1                                True                                True   

   urgent_care plan_name      pdf_name  
0          NaN   ABCBCBC  adjnajdn.pdf  
1          1.0   ABCBCBC  adjnajdn.pdf