RE-ORDER VALUES OF COLUMNS PARAMETERS

Question

I’m using this code valores.pivot(index='MARCA', columns='MES_C', values=['CMC','CMV','VL_VENDAS'])
to pivot this dataframe on python withe pandas.

MARCA MES_C CMC CMV VL_VENDAS
3F 01/2022 0,00 33,85 147,70
3M 01/2022 57.130,75 77.457,69 182.964,37
3M 02/2022 87.177,66 75.491,39 169.560,01

but the columns='MES_C' is out of order.

               CMC        CMV  VL_VENDAS        CMC         CMV   VL_VENDAS
MES_C      01/2022    01/2022    01/2022    02/2022     02/2022     02/2022
MARCA                                                                      
3F            0,00        NaN      33,85        NaN      147,70         NaN
3M        5.130,75   7.177,66   7.457,69   5.491,39    2.964,37    9.560,01

how to change to get the result like this?

               CMC                   CMV              VL_VENDAS            
MES_C      01/2022    02/2022    01/2022    02/2022     01/2022     02/2022
MARCA                                                                      
3F            0,00        NaN      33,85        NaN      147,70         NaN
3M       57.130,75  87.177,66  77.457,69  75.491,39  182.964,37  169.560,01

I want to show the results groped by the values [‘CMC’,’CMV’,’VL_VENDAS’]

I’ved tried with

valores.pivot(index='MARCA', columns='MES_C', values=['CMC','CMV','VL_VENDAS'])

and

pd.pivot_table(valores, values=['CMC','CMV','VL_VENDAS'], index='MARCA', columns='MES_C', aggfunc='first')

Asked By: maycondouglasmartinsrodrigues

||

Source

Answer 1

pivot = pd.pivot_table(valores, index=['MARCA'], columns='MES_C', aggfunc={'CMC':'sum','CMV':'sum','VL_VENDAS':'sum'}, fill_value=0, margins=True)
.swaplevel(axis=1)
.sort_index(level=0, axis=1)
.reindex(['CMC','CMV','VL_VENDAS'], level=1, axis=1)
.rename_axis(columns=[None, None])

pivot.to_excel('relatorio.xlsx', sheet_name='cmcxcmv')

with this code, works to change the column order and fix the problem.

Answered By: maycondouglasmartinsrodrigues

RE-ORDER VALUES OF COLUMNS PARAMETERS

Question:

Answers: