R：将类型“List”转换为 dataframe 以转换为 excel - 文本挖掘

Question

When trying to stem and tokenize my list of reviews, it will automatically make it a list. It is a "character" type variable at first, but when applying the following code it turns it into a "list":

reviews <- tokenize_word_stems(reviews)

I want to eventually convert this into excel, but my write_xlsx function can only convert dataframes, and not lists.

the rest of my code looks like this, but it goes "wrong" when trying to stem the words:

reviews <- readLines("Reviewlist.csv")
reviews <- gsub(pattern = "\\W", replace = " ", reviews)
reviews <- tolower(reviews)
reviews <- gsub(pattern="\\b[A-z]\\b{1}", replace=" ", reviews)
reviews <- stripWhitespace(reviews)
reviews <- removeWords(reviews, stopwords())
reviews <- tokenize_word_stems(reviews)

the file:

Thanks in advance!

Answer 1

Creating a lorem-ipsum dummy input here, based on my assumption what your "Reviewlist.csv" seems to look like.

library(dplyr)
library(stringi)

stri_rand_lipsum(5) %>%
  writeLines("Reviewlist.csv")

Then, this here is just your original code without alterations, but using dplyr grammar and explicitly stating the libraries necessary:

library(tm)
library(tokenizers)

reviews <- readLines("Reviewlist.csv") %>%
  gsub(pattern = "\\W", replace = " ", .) %>%
  tolower() %>%
  gsub(pattern="\\b[A-z]\\b{1}", replace=" ", .) %>%
  stripWhitespace() %>%
  removeWords(stopwords()) %>%
  tokenize_word_stems()

Now, what you can do is to bind your list items into a dataframe before being able to write it as an xlsx-file:

library(purrr)
library(writexl)

reviews_df <- reviews %>%
    map_dfr(~ setNames(., sprintf("word_%04d", seq_along(.))))

reviews_df %>%
  write_xlsx("Reviewlist.xlsx")

And that might create a very wide xlsx for you.
No idea whether Excel really is able to open it, but there you go:)