Pandas 数据框将行转换为列

Question

I've been trying to do this for a while and I've even checked out some other threads here but it seems like nobody's been having a problem like this until now. This is my dataframe

So, I've been trying to turn the "TipoPromo" column that contains 10 different possible values into 10 different columns, the values on these columns should be the ones that appear on the column "Desconto". Doing this I expect to reduce the number of rows on my dataframe significantly, and thus reducing the duplicates that you guys can see on the other rows (If you look at the columns "tickets" etc you'll see that there are many values that repeat themselves). I've tried using the pivot function but without much success. Could you guys understand what I'm trying to say?

Any help will be really appreciated!

Answer 1

pivot tables in pandas can solve this issue. I copied the top 5 and one below the ellipsis rows as a csv file, then:

df=pd.read_csv('data.txt',header=0)

cols=df.columns
cols=cols.drop(['Desconto','TipoPromo'])

df.pivot_table(columns='TipoPromo' , values='Desconto', index=cols.to_list()).reset_index()

This gives:

TipoPromo           Mes        Dias Semana DiaSemana       Bandeira           Categoria  Tickets  ...  Venda Liquida          CMV  CODIGO 12   DIGITAL  MARK DOWN (CODIGO 5)  MEU DESCONTO  NIVEL LOJA
0          JANEIRO/2019  2019-01-06    S01       DOM  PAO DE ACUCAR  PEREC COMPLEMENTER   152240  ...    1956066.000  1340467.172        NaN       NaN                   NaN      26463.40         NaN
1          OUTOBRO/2021  2021-10-25    S44       SEG    EXTRA HIPER             LIQUIDA    24344  ...     453297.096   376097.323        0.0  14363.72                -21.05       1852.07    5776.574