Removing specific words from word cloud in R

Question

I have made a word cloud in R for 2 songs. Now in the tdm when I display the items, i get the frequency of words for song 1 and song 2. I am also able to print the word cloud perfectly. My problem is i do not want words in tdm who frequency is less than 2. How can I do that.

I wrote the code and got this output:

tdm=TermDocumentMatrix(corpus)

> tdm=as.matrix(tdm)
>
> tdm
>

song 1  song 2
act                   0  2
action                0  2
actions               0  1
activity              5  4

I only want word activity as it occur more than once in both the songs. I mean I want to remove the words, act, action, actions. How can I do that ?

Answer 1

You didn't provide data some something like this should work:

data("crude")
tdm <- TermDocumentMatrix(crude)

x <- as.matrix(tdm)[, 1:2]
x[rowSums(apply(x, 2, ">", 1)) == 2, ]

Explanation: The line x <- as.matrix(tdm)[, 1:2] just getting 2 columns like your data so it doesn't do anything but needed to make data that looked like yours since you didn't provide any. This line apply(x, 2, ">", 1) says give me logical values for the statement is this greater than 1. Then I wrap this with rowSums (logical values are TRUE=1 and FALSE=0). Values equal to 2 (I had > 1 before but this is sloppy) are the conditions you're looking for. The I use a logical index with this output x[GRAB_THE_ROWS, ] . You can tear each step apart and run the code for yourself as seen below:

(step_1 <- apply(x, 2, ">", 1))
(step_2 <- rowSums(step_1))
(step_3 <- step_2 == 2)
x[step_3, ]