r - DocumentTermMatrix 需要有词频加权误差

Question

我正在尝试在相当大的数据集上使用 topicmodels 包中的 LDA()。在尝试了所有方法来修复以下错误“In nr * nc : NAs produced by integer overflow”和“输入矩阵的每一行需要包含至少一个非零条目”之后，我最终遇到了这个错误。

ask<- read.csv('askreddit201508.csv', stringsAsFactors = F)    
myDtm <- create_matrix(as.vector(ask$title), language="english", removeNumbers=TRUE, stemWords=TRUE, weighting=weightTf)
myDtm2 = removeSparseTerms(myDtm,0.99999)
myDtm2 <- rollup(myDtm2, 2, na.rm=TRUE, FUN = sum)
rowTotals <- apply(myDtm2 , 1, sum)
myDtm2   <- myDtm2[rowTotals> 0, ]  
LDA2 <- LDA(myDtm2,100)

Error in LDA(myDtm2, 100) : 
  The DocumentTermMatrix needs to have a term frequency weighting

Answer 1

问题的一部分是您通过 tf-idf 对文档术语矩阵进行加权，但 LDA 需要术语 counts。此外，这种删除稀疏术语的方法似乎正在创建一些已删除所有术语的文档。

使用 quanteda 包可以更轻松地将文本转换为主题模型。方法如下:

require(quanteda)
myCorpus <- corpus(textfile("http://homepage.stat.uiowa.edu/~thanhtran/askreddit201508.csv",
                            textField = "title"))
myDfm <- dfm(myCorpus, stem = TRUE)
## Creating a dfm from a corpus ...
##    ... lowercasing
##    ... tokenizing
##    ... indexing documents: 160,707 documents
##    ... indexing features: 39,505 feature types
##    ... stemming features (English), trimmed 12563 feature variants
##    ... created a 160707 x 26942 sparse dfm
##    ... complete. 

# remove infrequent terms: see http://stats.stackexchange.com/questions/160539/is-this-interpretation-of-sparsity-accurate/160599#160599
sparsityThreshold <- round(ndoc(myDfm) * (1 - 0.99999))
myDfm2 <- trim(myDfm, minDoc = sparsityThreshold)
## Features occurring in fewer than 1.60707 documents: 12579
nfeature(myDfm2)
## [1] 14363

# fit the LDA model
require(topicmodels)
LDA2 <- LDA(quantedaformat2dtm(myDfm2), 100)