Question

我有什么：

a <- c(49, 55, '2017-11-01', '13:54:57', '2017-11-01', '14:13:46')
b <- c(173, 171, '2017-11-01', '14:16:19', '2017-11-01', '14:23:34')
c <- c(171, 173, '2017-11-01', '14:57:01', '2017-11-01', '15:05:11')
d <- c(55, 49, '2017-11-01', '15:07:07', '2017-11-01', '15:27:30')
e <- as.data.frame(t(data.frame(a, b, c, d)))
colnames(e) <- c('start', 'stop', 'fdate', 'ftime', 'tdate', 'ttime')
e

  start stop      fdate    ftime      tdate    ttime 
     49   55 2017-11-01 13:54:57 2017-11-01 14:13:46
    173  171 2017-11-01 14:16:19 2017-11-01 14:23:34
    171  173 2017-11-01 14:57:01 2017-11-01 15:05:11
     55   49 2017-11-01 15:07:07 2017-11-01 15:27:30

我想在下一个条件中合并行：如果前一行ftime和ttime之间的差异小于30分钟，我们必须写一行：

第一行

start，最后一行stop，第一行ftime，最后一行ttime

我想要的例子：

  start stop      fdate    ftime      tdate    ttime
     49  171 2017-11-01 13:54:57 2017-11-01 14:23:34
    171   49 2017-11-01 14:57:01 2017-11-01 15:27:30

Answer 1

尝试以下

library(lubridate)
library(dplyr)

myfun <- function(e) {
    dt_f <- ymd_hms(paste(e$fdate, e$ftime, sep=" "))    # convert to Date
    lag_dt_t <- ymd_hms(paste(dplyr::lag(e$tdate, 1, default=head(e$fdate, 1)), dplyr::lag(e$ttime, 1, default=NA), sep=" "))   # convert to Date
    lgl <- dt_f - lag_dt_t <= make_difftime(minute = 30)   # is difference <= 30 min?
    lgl[1] <- FALSE       # Always set first row to FALSE since there is no previous row to compare to
    return(cumsum(!lgl))    # Returns `grp` values
}

temp <- e
temp$grp <- myfun(temp)
ans <- temp %>%
        group_by(grp) %>%
        summarise(  start = head(start, 1),
                stop = tail(stop, 1),
                fdate = head(fdate, 1),
                ftime = head(ftime, 1),
                tdate = tail(tdate, 1),
                ttime = tail(ttime, 1)) %>%
        ungroup() %>%
        select(-grp)

# A tibble: 2 x 6
   # start   stop      fdate    ftime      tdate    ttime
  # <fctr> <fctr>     <fctr>   <fctr>     <fctr>   <fctr>
# 1     49    171 2017-11-01 13:54:57 2017-11-01 14:23:34
# 2    171     49 2017-11-01 14:57:01 2017-11-01 15:27:30

Answer 2

这是另一个基本R替代方案：

UIView

使用dplyr计算各列之间各种值之间的差异

2 个答案: