通过前两个字符串对特定行进行子集

Question

我有一个数据框：

ID        value
1    package ‘glue’ was built under R version 3.6.2
2    Attaching package: ‘data.table’
3    Attaching package: ‘purrr’
4    Attaching package: ‘dplyr’
5    The following object is masked from ‘package:dplyr’
6    package ‘dplyr’ was built under R version 3.6.2 

如果值以附加包（第 2、3、4 行）开头，我想在第二个单词之后剪切其余部分并保持唯一。所以，想要的结果是：

ID        value
1    package ‘glue’ was built under R version 3.6.2
2    Attaching package:
5    The following object is masked from ‘package:dplyr’
6    package ‘dplyr’ was built under R version 3.6.2 

我怎么能这样做。我知道我应该使用正则表达式：

\\S+\\s+\\S+

但是剩下的怎么做呢？

Answer 1

df$value <- sub("(Attaching package:)(.*)", "\\1", df$value)

您可以将要省略的部分定义为捕获组，不在subs 替换参数中引用它，从而有效地删除它。

要删除 dups，请使用 duplicated:

df <- df[!duplicated(df$value),] 

Answer 2

我使用 rebus 包提供了一个解决方案：

library(dplyr)
library(rebus)
library(stringi)

df <- dplyr::tibble(ID = c(1:6),
                    value =c("package ‘glue’ was built under R version 3.6.2","Attaching package: ‘data.table’","Attaching package: ‘purrr’","Attaching package: ‘dplyr’","The following object is masked from ‘package:dplyr’","package ‘dplyr’ was built under R version 3.6.2" ))

pattern <- START %R% "Attaching package:"

df %>% 
  dplyr::mutate(NEW = ifelse(stringi::stri_detect_regex(value, pattern), "Attaching package", value)) %>% 
  dplyr::distinct(NEW)

  NEW                                                
  <chr>                                              
1 package ‘glue’ was built under R version 3.6.2     
2 Attaching package                                  
3 The following object is masked from ‘package:dplyr’
4 package ‘dplyr’ was built under R version 3.6.2 

Answer 3

这行得通吗：

library(dplyr)
df %>% mutate(value = gsub('(^Attaching package:).*', '\\1',value)) %>% 
group_by(value) %>% filter(ID == min(ID))
# A tibble: 4 x 2
# Groups:   value [4]
     ID value                                              
  <dbl> <chr>                                              
1     1 package ‘glue’ was built under R version 3.6.2     
2     2 Attaching package:                                 
3     5 The following object is masked from ‘package:dplyr’
4     6 package ‘dplyr’ was built under R version 3.6.2    
 

使用 stringr、str_remove、str_replace 并往后看：

library(stringr)
df %>% mutate(value = str_remove(value, '(?<=Attaching package:).*')) %>% group_by(value) %>% filter(ID == min(ID))
# A tibble: 4 x 2
# Groups:   value [4]
     ID value                                              
  <dbl> <chr>                                              
1     1 package ‘glue’ was built under R version 3.6.2     
2     2 Attaching package:                                 
3     5 The following object is masked from ‘package:dplyr’
4     6 package ‘dplyr’ was built under R version 3.6.2  

df %>% mutate(value = str_replace(value, '(^Attaching package:).*', '\\1')) %>% group_by(value) %>% filter(ID == min(ID))
# A tibble: 4 x 2
# Groups:   value [4]
     ID value                                              
  <dbl> <chr>                                              
1     1 package ‘glue’ was built under R version 3.6.2     
2     2 Attaching package:                                 
3     5 The following object is masked from ‘package:dplyr’
4     6 package ‘dplyr’ was built under R version 3.6.2