如何在分组列中按行添加值答案

【问题标题】：How to add values rowwise in a grouped column如何在分组列中按行添加值
【发布时间】：2018-07-24 15:29:03
【问题描述】：

我有一些传感器数据，每秒有 100 个数据条目。最后一列是毫秒，现在都是 10。我怎样才能将毫秒按行相加，按时间和日期分组。

testdata <- structure(list(local_date = c("26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017", "26-06-2017",  "26-06-2017", "26-06-2017"), 
                           local_time = c("13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23", "13:58:23",  "13:58:23", "13:58:23", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24",  "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24", "13:58:24" ), 
                           ms = c(10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,  10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10)), 
                      .Names = c("local_date",  "local_time", "ms"), row.names = c(NA, -200L), class = c("data.table", "data.frame"))

前 100 行都共享相同的时间 (13:58:23) 和日期 (26-06-2017)，但它们都有 10 毫秒。结果应该只有一个每秒 10 毫秒的条目，然后将后面的毫秒添加到前面的毫秒中。

这个 sn-p 将创建一个序列的结果：

testdata$ms = rep(seq(from = 10, to = 1000, by = 10), 2)

但由于原始数据不是那么干净，我必须按日期和时间对数据进行分组，然后以行方式将毫秒相加。

我更喜欢data.table 解决方案，但dplyr 也可以正常工作。

【问题讨论】：

标签： r data.table rowwise

【解决方案1】：

听起来你需要一个分组的cumsum：

library(dplyr) 

testdata$ms2 = rep(seq(from = 10, to = 1000, by = 10), 2)

testdata %>%
    group_by(local_date, local_time) %>%
    mutate(cumsum_ms = cumsum(ms))

   local_date local_time    ms   ms2 cumsum_ms
   <chr>      <chr>      <dbl> <dbl>     <dbl>
 1 26-06-2017 13:58:23      10    10        10
 2 26-06-2017 13:58:23      10    20        20
 3 26-06-2017 13:58:23      10    30        30
 4 26-06-2017 13:58:23      10    40        40
 5 26-06-2017 13:58:23      10    50        50

【讨论】：

【解决方案2】：

并添加一个data.table 版本：

testdata[, ms := cumsum(ms), by = .(local_time, local_date)]

【讨论】：