python - 如何根据重复值汇总列？

Question

我正在尝试使用带有条件的 for 循环来计算事件的持续时间。

mydf= {'Duration' : [14, 8, 6, 36, 12, 5, 3, 2, 4, 5, 8, 3, 14, 1, 27, 25, 117, 2, 962, 2, 2, 1], 
'Activity': ['Groom', 'Pause', 'Groom', 'Sleep', 'Sleep', 'Sleep', 'Sleep', 'Sleep', 'Sleep', 'Sleep', 'Sleep', 'Sleep', 'Sleep', 'Awaken', 'Groom', 'Pause', 'Groom', 'Eat', 'Cuddle', 'Come down', 'Dig', 'Forage']}
    df = pd.DataFrame(mydf)

我想将所有 sleep 持续时间加在一起。

我尝试过:

Sleep_sum = [sum(df['Duration'] for i in df['Activity'] if [i+1]=='Sleep')] 

但这给了我一个错误:TypeError: can only concatenate str (not "int") to str.

我也尝试过这个:

for i in range of len(df):
     df['Activity'][i] == 'Sleep'
     if [i] = [i+1]
     df['Duration'].sum()

本质上，如果后面跟着另一行==“ sleep ”，我需要对“ sleep ”的持续时间进行求和。

感谢您的宝贵时间!

Answer 1

您可以创建虚拟组来对连续事件( sleep 和其他)进行分组:

out = (df.groupby(df['Activity'].ne(df['Activity'].shift()).cumsum(), as_index=False)
         .agg({'Duration': 'sum', 'Activity': 'first'}))
print(out)

# Output
    Duration   Activity
0         14      Groom
1          8      Pause
2          6      Groom
3         92      Sleep
4          1     Awaken
5         27      Groom
6         25      Pause
7        117      Groom
8          2        Eat
9        962     Cuddle
10         2  Come down
11         2        Dig
12         1     Forage

更新

仅用于 sleep 事件:

m = df['Activity'] == 'Sleep'
sleep = (df.reset_index(drop=False)[m]
           .groupby(df['Activity'].ne(df['Activity'].shift()).cumsum())
           .agg({'index': 'first', 'Duration': 'sum', 'Activity': 'first'})
           .set_index('index'))

out = pd.concat([df[~m], sleep]).sort_index()
print(out)

# Output
    Duration   Activity
0         14      Groom
1          8      Pause
2          6      Groom
3         92      Sleep
13         1     Awaken
14        27      Groom
15        25      Pause
16       117      Groom
17         2        Eat
18       962     Cuddle
19         2  Come down
20         2        Dig
21         1     Forage