如何在 R 中按组填充数据框中的空白？

Question

我有一个包含时间序列、一组事件和一个 ID 的数据框。我通过共享时间列连接了两个数据框，但我想填补事件中的空白。

这是一个简化的示例，其中包含我用来连接数据框的值列，然后是事件和 ID 列。我想以某种方式在第一个（开始）和最后一个（结束）事件条目之间填写事件和 ID 列。


a<-c(seq(from = 150, to = 213, by = 3))
b<-c("start","","","mid","","end","", "",
         "start", "", "", "end", "", "", "",
         "start", "", "end", "start", "mid", "", "end")
c<-c("A","","","A","","A","", "",
         "A", "", "", "A", "", "", "",
         "B", "", "B", "B", "B", "", "B")

(data<-data.frame(value = a, event = b, ID = c))

这里是目标，开始和结束之间填写了事件和ID：

agoal<-c(seq(from = 150, to = 213, by = 3))
bgoal<-c("start","start","start","mid","mid","end","", "",
         "start", "start", "start", "end", "", "", "",
         "start", "start", "end", "start", "mid", "mid", "end")
cgoal<-c("A","A","A","A","A","A","", "",
         "A", "A", "A", "A", "", "", "",
         "B", "B", "B", "B", "B", "B", "B")

(goal<-data.frame(value = agoal, event = bgoal, ID = cgoal))

Answer 1

您可以使用 dplyr 和 tidyr 完成此任务：

library(tidyr)
library(dplyr)

data %>% 
  mutate(grp = cumsum(case_when(event == "end" ~ -1,
                         event == "start" ~ 1,
                         TRUE ~ 0)),
         across(c(-value, -grp), ~ ifelse(.x == "" & grp == 1, NA_character_, .x))) %>% 
  fill(c(-value), .direction="down") %>% 
  select(-grp)

returns

   value event ID
1    150 start  A
2    153 start  A
3    156 start  A
4    159   mid  A
5    162   mid  A
6    165   end  A
7    168         
8    171         
9    174 start  A
10   177 start  A
11   180 start  A
12   183   end  A
13   186         
14   189         
15   192         
16   195 start  B
17   198 start  B
18   201   end  B
19   204 start  B
20   207   mid  B
21   210   mid  B
22   213   end  B

Answer 2

这是另一个选项data.table

将 data.frame 转换为 data.table - setDT
遍历 .SDcols - 'event'、'ID'
将空格 ("") 替换为 NA - na_if
使用 na.locf0（来自 zoo）用之前的非 NA 填充 NA 元素并将（:=）分配回列
获取行索引 (.I)，其中 'event' 值为 duplicated，其中 'event' 为按运行- 分组的“结束”事件的长度 ID (rleid)
提取行索引（$V1）并将那些'event'、'ID'分配给空白

library(data.table)
library(zoo)
library(dplyr)
setDT(data)[, c("event", "ID") := lapply(.SD, function(x) 
        na.locf0(na_if(x, ""))), .SDcols = event:ID]
data[data[, .I[duplicated(event) & event == "end"] , 
         rleid(event)]$V1, c("event", "ID") := .("", "")]

-输出

data
    value event ID
 1:   150 start  A
 2:   153 start  A
 3:   156 start  A
 4:   159   mid  A
 5:   162   mid  A
 6:   165   end  A
 7:   168         
 8:   171         
 9:   174 start  A
10:   177 start  A
11:   180 start  A
12:   183   end  A
13:   186         
14:   189         
15:   192         
16:   195 start  B
17:   198 start  B
18:   201   end  B
19:   204 start  B
20:   207   mid  B
21:   210   mid  B
22:   213   end  B

如何在 R 中按组填充数据框中的空白？

How can I fill in gaps in a data frame by group in R?

r

grouping

sequence

series

fill