问题解决了：
R -
data = data %>% group_by(ID) %>% mutate(
  ranks = rank(index)
)
python -
a['ranks'] = a.groupby(['user_id'])['time'].rank(ascending=False,method='first')