数据框的基于行的过滤器

Question

I want to perform analysis on a dataframe. This is my dataframe format.

df_Input = pd.read_excel("/home/cc/Downloads/date.xlsx")

ID    | BOOK |  Type
-----------------------     
1     | ABC  |   MAR
45    | PQR  |   TAB
45    | EDF  |   Fin
1     | DCF  |   oop
45    | PQR  |   TAB

I want to find count(count of every unique value) and unique values that each unique ID can hold. The output should be a dataframe as shown below.

ID  |  BOOK_Count | Book_values  |Type_count |  Type_values
-----------------------------------------------------------
1   |    2        |  [ABC,DCF]   | 1         |    [MAR,oop]
45  |    2        |  [PQR,EDF]   | 2         |    [Fin,TAB]

I tried doing it but with a lot of loops. Thanks in advance

Answer 1

IIUC, you can use this:

df_out = df.groupby('ID')['BOOK','Type'].agg(['nunique', lambda x: list(set(x))])
df_out = df_out.rename(columns={'nunique':'count', '<lambda>':'values'})
df_out.columns = df_out.columns.map('_'.join)
print(df_out)

OUtput:

      BOOK_count BOOK_values  Type_count Type_values
ID                                                  
1_1            2  [ABC, DCF]           2  [MAR, oop]
45_2           2  [EDF, PQR]           2  [TAB, Fin]

Answer 2

Let's say we have this dataframe :

    ID  BOOK type
0   1   ABC  MAR
1   0   PQR  TAB
2   1   EDF  Fin
3   0   DCF  oop
4   1   PQR  TAB

You can use json aggregate format as follow :

aggreg = {
'BOOK':{
    'BOOK_COUNT' : len,
    'BOOK_values' : lambda r : r.tolist()
},

'type':{
    'Type_COUNT' : len,
    'Type_values' : lambda r : r.tolist()
} 
}

Then, use groupby :

df.groupby('ID').agg(aggreg)

#output :
          BOOK              type
    BOOK_COUNT  BOOK_values Type_COUNT  Type_values
ID              
0            2    [PQR, DCF]        2   [TAB, oop]
1            3    [ABC, EDF, PQR]   3   [MAR, Fin, TAB]