提取字符串中的所有单词和字母簇，然后使用 R 中的 gsub() 使每个单词成为一个单独的数据

Question

Say we have:

stringTest <- c("Here we have 4 words", "Here we have avwerfaf 4")

Expected output:

"Here" "we" "have" "words" "Here" "we" "have" "avwerfaf"

I would like to use gsub(), but other methods are definitely excepted. Thanks Guys!

Answer 1

You can use strsplit :

result <- unlist(strsplit(stringTest, " |\\d"))
result[result != ""]
#> [1] "Here"     "we"       "have"     "words"    "Here"     "we"      
#> [7] "have"     "avwerfaf"

or if you prefer a one-liner:

unlist(lapply(strsplit(stringTest, "\\W|\\d"), function(x) x[x != ""]))

Answer 2

library(tidyverse)

stringTest <- c("Here we have 4 words", "Here we have avwerfaf 4")

gsub(" \\d", replacement = "", stringTest) %>%
  str_split(pattern = " ") %>%
  unlist()

Answer 3

This falls under the "another approach" category. What you appear to be doing is tokenizing by words, dropping numbers.

library(tokenizers)

unlist(tokenize_words(stringTest, lowercase = FALSE, strip_numeric = TRUE))

Which gives:

[1] "Here"     "we"       "have"     "words"    "Here"     "we"       "have"     "avwerfaf"

---

If you are operating out of a data frame, something like this could be useful.

library(dplyr)
library(tidytext)

df <- tibble(description = stringTest)

df2 <- df %>% 
  rowid_to_column() %>% 
  unnest_tokens(word, description, to_lower = FALSE, strip_numeric = TRUE)

Which returns a new tibble:

> df2
# A tibble: 8 x 2
  rowid word    
  <int> <chr>   
1     1 Here    
2     1 we      
3     1 have    
4     1 words   
5     2 Here    
6     2 we      
7     2 have    
8     2 avwerfaf