Question

我正在R中使用tm包来处理波斯语文本，并且我正在使用主题建模，我有此代码，但在某些级别上它会向我发送错误代码是

##### load 'tm' package #####
library(tm)


##### load 'tm' package #####
library(tm)

##### read news files and convert them into a dataframe #####
newsFiles_dir <-"./data"
newsFiles<-list.files(newsFiles_dir,pattern = "*.txt",full.names=TRUE)

txt_data<- lapply(newsFiles,function( x ){strs<-read.delim(x,stringsAsFactors = FALSE)
paste(strs , collapse = ' ')})
data<-  data.frame( doc_id = newsFiles ,  text = as.character( unlist( do.call(rbind , txt_data))))

##### remove punctuations and control characters #####
data[[2]]<- gsub("ها" ," " ,data[[2]])
data[[2]]<-gsub("'","",data[[2]])
data[[2]]<- gsub("[[:punct:]]","",data[[2]])
data[[2]]<-gsub("[[:cntrl:]]", " ",data[[2]])
data[[2]]<-gsub("^[[:space:]]+", "",data[[2]])
data[[2]]<-gsub("[[:space:]]+$", "",data[[2]])

##### create corpus object #####
docs <- Corpus(VectorSource(data[[2]]), readerControl = list(language="fa_FAE",encoding ="UTF-8"))

##### remove stop words #####
persianStopwords_file_loc ="./stopwords/stopwords.txt"
persianStopwords<-readLines(persianStopwords_file_loc )

docs<-tm_map(docs,removeWords,persianStopwords)

##### create doctment-term matrix #####
dm<-DocumentTermMatrix(docs,control = list(encoding = 'UTF-8'))
rownames(dm)<-data[[1]]

##### remove empty documents , if any #####
rowTotals <- apply(dm,1,sum)
dtm <- dm [rowTotals>0, ]


##### load topicmodels library #####
library(topicmodels)

#model paramaters
burnin = 1000
iter = 1000 
keep = 50 
k = 10

##### train topic models ####
model <- LDA(dtm, k=k,method="Gibbs",control=list(burnin=burnin,iter=iter,keep=keep))
#model<-LDA(dtm,k=k)


##### topics assigned to each document #####
model.topics<-as.matrix(topics(model))
model.topics

##### how many documents in each topic #####
table(model.topics)

##### print top 5 important words in each topic #####
print(terms(model,5))


##### function to plot wordcloud for a given topic #####
plotWordCloud<- function(t){
  m<- as.matrix(dtm[rownames(model.topics)[which(model.topics==t)], ])
  require(wordcloud)
  v<- sort(colSums(m) , decreasing = TRUE)
  words<- names(v)
  d<- data.frame(words= words , frq=v)
  dark2<- brewer.pal(6,"Dark2")
  wordcloud(d$words , d$frq , min.freq = 2 , colors = dark2)

}

##### plot wordcloud of two topics and compare them! #####
par(mfrow=c(1,2))

plotWordCloud(5)
plotWordCloud(10)

，错误是错误（函数（类，fdef，mtable）：无法找到签名““ list””的“ topics”功能的继承方法，有人吗？

R中的波斯文本挖掘

0 个答案: