0

I Have a dataframe wich has this structure : 

Note.Reco Review Review.clean.lower
10 Good Products  good products
9 Nice film      nice film
....         ....

The first column is the rank of the film, then the second column is the custmer's review then the 3rd column is the review with lowercase letters.

I try now to delete stop words with this :

Data_clean$Raison.Reco.clean1 <- Corpus(VectorSource(Data_clean$Review.clean.lower))
Data_clean$Review.clean.lower1 <- tm_map(Data_clean$Review.clean.lower1, removeWords, stopwords("english"))

But R studio crashes

Can you help me to resolve this problem please?

Thank you

EDIT : 

#clean up
# remove grammar/punctuation
Data_clean$Review.clean.lower <- tolower(gsub('[[:punct:]0-9]', ' ', Data_clean$Review))

Data_corpus <- Corpus(VectorSource(Data_clean$Review.clean.lower))

Data_clean <- tm_map(Data_corpus,  removeWords, stopwords("french"))

train <- Data_clean[train.index, ]
test <- Data_clean[test.index, ]

So I get error when I run the 2 last instructions .

dr.nasri84
  • 79
  • 2
  • 9
  • To subset a corpus use `Data_clean[train.index]`. A corpus is basically a list so use list subsetting not matrix/dataframe subsetting – emilliman5 May 23 '17 at 12:58

1 Answers1

1

Try the below . You can do cleaning on the corpus and not column directly. 

Data_corpus <-
  Corpus(VectorSource(Data_clean$Review.clean.lower))

  Data_clean <- tm_map(Data_corpus,  removeWords, stopwords("english"))

EDIT: As mentioned by you, you want to be able to access the output after removing stop words, try the below instead of the above:

library(tm)

stopWords <- stopwords("en")

Data_clean$Review.clean.lower<- as.character(Data_clean$Review.clean.lower)
 '%nin%' <- Negate('%in%')
 Data_clean$Review.clean.lower1<-lapply(Data_clean$Review.clean.lower, function(x) {
  chk <- unlist(strsplit(x," "))
  p <- chk[chk %nin% stopWords]
  paste(p,collapse = " ")
})

Sample Output of above code:

>  print(Data_clean)
>       note Note.Reco.Review Review.clean.lower Review.clean.lower1
>     1   10    Good Products      good products       good products
>     2    9        Nice film     is a nice film           nice film

Also check the below: R remove stopwords from a character vector using %in%

MysticRenge
  • 373
  • 1
  • 4
  • 13