fork download
  1. import pandas as pd
  2. df = pd.DataFrame({'user': [1, 2,
  3. 3, 4],
  4. 'occupation': [380, 380, 24, 26]})
  5.  
  6. dic = df.groupby('user')['occupation'].apply(list).to_dict()
  7.  
  8. # for d in dic:
  9. # print(d, dic[d][0])
  10.  
  11. keys = df['occupation'].unique().astype(int)
  12. idx = {key: [] for key in keys}
  13.  
  14.  
  15. for i in range(1, len(dic)+1):
  16. occ = int(dic[i][0])
  17. idx[occ].append(i)
  18.  
  19. for i in idx:
  20. print(i, idx[i])
Success #stdin #stdout 0.31s 60288KB
stdin
Standard input is empty
stdout
(24, [3])
(26, [4])
(380, [1, 2])