根据后续行的内容有条件地替换值

Question

我有一个按时间顺序列出人们日常活动的数据集。一天中的第一项活动(从午夜开始)被称为“开始”，通常是在醒来之前，但并不总是如此。对于每个人(group_by(id))，我想把“开始”改为“睡眠”，如果“开始”先于“觉醒”。但是如果“开始”并不先于“觉醒”，至于id 22，那么我想保留“开始”。

id <- c(11,11,11,11,22,22,22,22,22,22,33,33,33)
activity <-c("Start","Wake","TV","Eat","Start","TV","Sleep","Wake","Eat","Dressed","Start","Wake","BrushTeeth")
DF<- data.frame(id,activity)
DF
   id   activity
1  11      Start
2  11       Wake
3  11         TV
4  11        Eat
5  22      Start
6  22         TV
7  22      Sleep
8  22       Wake
9  22        Eat
10 22    Dressed
11 33      Start
12 33       Wake
13 33 BrushTeeth

这就是我希望最终数据看起来的样子(请注意，第1行和第7行中的“开始”已被“睡眠”所取代，但在第5行中仍然是“开始”，因为它不先于"Wake")

   id   activity
1  11      Sleep
2  11       Wake
3  11         TV
4  11        Eat
5  22      Start
6  22         TV
7  22      Sleep
8  22       Wake
9  22        Eat
10 22    Dressed
11 33      Sleep
12 33       Wake
13 33 BrushTeeth

Answer 1

试试看

library(dplyr)

 DF %>% 
   mutate(new = replace(activity, activity == 'Start' & lead(activity) == 'Wake', 'Sleep'))

这给了，

id activity new 1 11 Start Sleep 2 11 Wake Wake 3 11 TV TV 4 11 Eat Eat 5 22 Start Start 6 22 TV TV 7 22 Sleep Sleep 8 22 Wake Wake 9 22 Eat Eat 10 22 Dressed Dressed 11 33 Start Sleep 12 33 Wake Wake 13 33 BrushTeeth BrushTeeth

Answer 2

data.table的一个选项是指定具有逻辑条件的i，并为那些从i返回为‘睡眠’的行指定(:=)活动

library(data.table)
setDT(DF)[activity == 'Start' & shift(activity, type = 'lead') == 'Wake',
        activity := 'Sleep'][]  

DF
#    id   activity
# 1: 11      Sleep
# 2: 11       Wake
# 3: 11         TV
# 4: 11        Eat
# 5: 22      Start
# 6: 22         TV
# 7: 22      Sleep
# 8: 22       Wake
# 9: 22        Eat
#10: 22    Dressed
#11: 33      Sleep
#12: 33       Wake
#13: 33 BrushTeeth