Načetní postů z facebooku, tokenizace, sestavení korpusu a vytvoření wordcloudu. Pro StuNoMe Digital Humanities

rfacebook-wordcloud.R

### Rfacebook

# Instalace knihovny - staci spustit jenom jednou
install.packages("Rfacebook")

# Nacteni knihovny
library(Rfacebook)

# Token ziskany z https://developers.facebook.com/tools/explorer
token <- "..."

## Uzivatelske udaje
# Nacteni udaju o sobe
me <- getUsers("me", token=token)
# Nacteni site vazeb do matice
network <- getNetwork(token, format="adj.matrix")

## Nacteni 100 prispevku stranky
stunome <- getPage("stunome", token=token, n=100)
summary(stunome$likes_count)

# Post s maximem likes
stunome[which.max(stunome$likes_count),]
# Posty s nadprumernym poctem likes
uspesne <- subset(stunome, likes_count >= mean(stunome$likes_count))


### Wordcloud

## Instalace zavislosti - staci spustit jenom jednou
install.packages("tm")
install.packages("wordcloud")

# Nacteni knihoven
library(tm)
library(wordcloud)

## Priprava korpusu
corpus <- Corpus(VectorSource(stunome$message))

corpus <- tm_map(corpus, tolower)
corpus <- tm_map(corpus, removePunctuation)
corpus <- tm_map(corpus, removeNumbers)

# Matice slov
ap.tdm <- TermDocumentMatrix(corpus)
ap.m <- as.matrix(ap.tdm)
# Dataframe frekvenci
ap.v <- sort(rowSums(ap.m),decreasing=TRUE)
ap.d <- data.frame(word = names(ap.v),freq=ap.v)

## Wordcloud
require(RColorBrewer)
# Sestaveni barevne palety
# pro vyber palet viz http://blog.ggplot2.org/post/24607351280/choosing-colour-palettes-part-ii-educated-choices
pal2 <- brewer.pal(8,"Dark2")

# Povolit zapis obrazku do souboru
png("wordcloud_stunome.png", width=1024,height=768)
# Sestaveni wordcloudu
wordcloud(ap.d$word,ap.d$freq, scale=c(10,.2),min.freq=3,
          max.words=150, random.order=FALSE, rot.per=.15, colors=pal2)
# Vypne zapis do souboru (device off)
dev.off()