Question

此玩具数据框表示人员输入的时间。我可以使用的格式有完全相同的模式，用于同一个人和同一天的多个文本条目。同一个人和同一天最多可以输入15个文本。多文本条目的行中没有人员条目。

structure(list(Date = structure(c(1514764800, 1514764800, NA, 
1517443200, 1519862400, NA, NA, NA, 1519862400, NA, NA), class = c("POSIXct", 
"POSIXt"), tzone = "UTC"), Person = c("FMC", "ABC", NA, "FMC", 
"ABC", NA, NA, NA, "RWM", NA, NA), Text = c("work on request", 
"More text", "third line", "email to re: summary", "work on loan documents", 
"sixth line of text", "text seven", "eighth in a series", "conferences with working group", 
"line ten", "review and provide comments")), row.names = c(NA, 
-11L), class = c("tbl_df", "tbl", "data.frame"))

如何组合文本元素，以便每个人的条目每天只有一行，删除不需要的行（一旦文本粘贴在一起）并到达以下对象？

编辑后的问题忽略了我尝试失败的for循环。

必须有一种方法可以将给定日期的所有文本合并为给定日期的一行（例如，ABC在2018年1月1日有两个条目）并删除合并后的行文字来了。

Answer 1

我们可以使用na.locf用最后一个非缺失值填充缺失值（NA），然后group_by连续出现Person和summarise通过Text一起paste来实现。

library(dplyr)
library(zoo)
library(data.table)

df %>%
  na.locf(.) %>%
  group_by(group = rleid(Person)) %>%
  summarise(Text = paste0(Text, collapse = " "))


#  group Text                                                                   
#  <int> <chr>                                                                  
#1     1 work on request                                                        
#2     2 More text third line                                                   
#3     3 email to re: summary                                                   
#4     4 work on loan documents sixth line of text text seven eighth in a series
#5     5 conferences with working group line ten review and provide comments

对于更新后的问题，我们可以

library(dplyr)
library(zoo)

df %>%
  na.locf(.) %>%
  group_by(Date, Person) %>%
  summarise(Text = paste0(Text, collapse = " "))

Answer 2

无需复杂，只需使用tidyverse。

针对问题的更改进行了调整：

library(tidyverse)

> df%>%
   fill(Date:Person, Date:Person) %>% # Fills missing values in using the previous entry.
   group_by(Date, Person) %>%
   summarise(Text = paste(Text, collapse = ' '))

# A tibble: 5 x 3
  Date                Person Text                                                                   
  <dttm>              <chr>  <chr>                                                                  
1 2018-01-01 00:00:00 ABC    More text third line                                                   
2 2018-01-01 00:00:00 FMC    work on request                                                        
3 2018-02-01 00:00:00 FMC    email to re: summary                                                   
4 2018-03-01 00:00:00 ABC    work on loan documents sixth line of text text seven eighth in a series
5 2018-03-01 00:00:00 RWM    conferences with working group line ten review and provide comments

数据：

# A tibble: 11 x 3
   Date                Person Text                          
   <dttm>              <chr>  <chr>                         
 1 2018-01-01 00:00:00 FMC    work on request               
 2 2018-01-01 00:00:00 ABC    More text                     
 3 NA                  NA     third line                    
 4 2018-02-01 00:00:00 FMC    email to re: summary          
 5 2018-03-01 00:00:00 ABC    work on loan documents        
 6 NA                  NA     sixth line of text            
 7 NA                  NA     text seven                    
 8 NA                  NA     eighth in a series            
 9 2018-03-01 00:00:00 RWM    conferences with working group
10 NA                  NA     line ten                      
11 NA                  NA     review and provide comments

Answer 3

library(dplyr)

merge_lines <- function(x) paste(x, collapse = ' ')

df %>% 
  zoo::na.locf(.) %>%
  group_by(Person) %>%
  summarise_at(vars(Text), (funs(merge_lines)))

结果：

# A tibble: 4 x 2
  Person Text                                                                   
  <chr>  <chr>                                                                  
1 ABC    More text third line                                                   
2 FMC    work on request email to re: summary                                   
3 HIL    work on loan documents sixth line of text text seven eighth in a series
4 RWM    conferences with working group line ten review and provide comments

在数据框中合并文本元素，并删除文本来自的行

3 个答案: