另一列不随时间变化

Question

我有这样一个数据帧：

unit start stop A 0.0 8.15 A 9.18 11.98 A 13.07 13.80 B 13.82 15.00 B 16.46 17.58 A 17.62 17.98 B 18.01 19.99 B 20.10 25.11

Answer 1

你实际上在找tally=cumcount, time=cumsum。所以我会这么做：

# these are the blocks
s = df['unit'].ne(df['unit'].shift()).cumsum()

# time for each row
times = df['stop'] - df['start']

# compute the new df 
new_df = (times.groupby(s)
            .agg(tally='cumcount', time='cumsum')  
            .assign(unit=df['unit'])               
            .pivot(columns='unit', 
                   values=['tally', 'time'])
            .fillna(0)
         )

# rename the columns
new_df.columns = [f'unit{y}_{x}' for x,y in new_df.columns]

# concat
pd.concat((df, new_df), axis=1)

输出：

  unit  start   stop  unitA_tally  unitB_tally  unitA_time  unitB_time
0    A   0.00   8.15          0.0          0.0        8.15        0.00
1    A   9.18  11.98          1.0          0.0       10.95        0.00
2    A  13.07  13.80          2.0          0.0       11.68        0.00
3    B  13.82  15.00          0.0          0.0        0.00        1.18
4    B  16.46  17.58          0.0          1.0        0.00        2.30
5    A  17.62  17.98          0.0          0.0        0.36        0.00
6    B  18.01  19.99          0.0          0.0        0.00        1.98
7    B  20.10  25.11          0.0          1.0        0.00        6.99