R convert dataframe to term-document-matrix

Question

I'm currently learning my ways around R and Im troubled by the following problem:

Ive got a dataframe that is build up like this

word       freq1        freq2

tree        10           20
this         2            3
that         4            5
...

It shows the frequency in which the word is used in text 1 (freq1) and text 2 (freq2). Is it possible to transform this to a term-document-matrix? I need it to be a term-document-matrix to apply the following function

par(mfrow=c(1,1))
comparison.cloud(tdm, random.order=FALSE, colors = 
c("indianred3","lightsteelblue3"),
title.size=2.5, max.words=400)

from https://rpubs.com/brandonkopp/creating-word-clouds-in-r

Thanks :)

Have you tried the `tm` library or if you already have the terms `wordcloud` and `wordcloud2`? — NelsonGon, Jan 23 '19 at 09:54
Hello Hunterofdark, maybe this package/function is helpful: https://www.rdocumentation.org/packages/qdap/versions/2.3.0/topics/as.tdm — TinglTanglBob, Jan 23 '19 at 10:13

NelsonGon · Accepted Answer · 2019-01-23T10:52:36.647

EDIT: After reshaping your data:

library(reshape2)
library(tm)
library(dplyr)
library(wordcloud)
df2<-df %>% 
  gather("Origin","Freq",c(2,3)) %>% 
  acast(word~Origin,fill=0,value.var = "Freq")
comparison.cloud(df2, random.order=FALSE, colors = c("indianred3","lightsteelblue3"),
                 max.words=400)

Result:

Original answer: There is something wrong with your data as it stands. Here is a basic workflow leading up to either a wordcloud or comparison cloud.

library(tm)
library(dplyr)
library(wordcloud)
df<-read.table(text="word       freq1        freq2

               Tree        10           20
               This         2            3
               That         4            5",header=T)
df$word<-as.character(df$word)
df1<-df %>% 
  gather()
corpus_my<-Corpus(VectorSource(df1))
tdm<-as.matrix(TermDocumentMatrix(corpus_my))
comparison.cloud(tdm, random.order=FALSE, colors = c("indianred3","lightsteelblue3"),
                 max.words=400)

This gives which is not what you expect. I would suggest restructuring your data first:

This looks helpful! I'll try out when I'm at home. I was aware that probably the structure of my data was a problem but I wasnt sure how to fix it. I'll come back to comment after I tried it out. Many thanks already for your effort :) — Hunterofdark91, Jan 23 '19 at 11:10

R convert dataframe to term-document-matrix

1 Answers1