图3-9-3 单篇文章的词云图.R有问题

[shixiangbupt]

commonality.cloud

library(tm)
library(wordcloud)

Paper1<-paste(scan("Paper1.txt", what = character(0),sep =""),collapse = " ") #读入 TXT 文档 1
Paper2<-paste(scan("Paper2.txt", what = character(0),sep =""),collapse = " ") #读入 TXT 文档 2
tmpText<- data.frame(c(Paper1,Paper2),row.names=c("Text1","Text2"))
df_title <- data.frame(doc_id=row.names(tmpText),text=tmpText$c.Paper1..Paper2.)
ds <- DataframeSource(df_title)
#创建一个数据框格式的数据源，首列是文档 id(doc_id),第二列是文档内容
corp = VCorpus(ds)
#加载文档集中的文本并生成语料库文件
corp = tm_map(corp,removePunctuation) #清除语料库内的标点符号
corp = tm_map(corp,PlainTextDocument) #转换为纯文本
corp = tm_map(corp,removeNumbers) #清除数字符号
corp = tm_map(corp,function(x){removeWords(x,stopwords())}) #过滤停止词库
term.matrix <- TermDocumentMatrix(corp)
#利用 TermDocumentMatrix()函数将处理后的语料库进行断字处理，生成词频权重矩阵
term.matrix <- as.matrix(term.matrix) #频率
colnames(term.matrix) <- c("Paper1","paper2")
#comparison.cloud(term.matrix,max.words=300,random.order=FALSE,colors=c("#00B2FF", "red"))
#图 3-9-4(a)
commonality.cloud(term.matrix,max.words=100,random.order=FALSE,color="#E7298A")
#图 3-9-4(b)
df<-data.frame(term.matrix)
#wordcloud(row.names(df) , df$Paper1 ,min.freq=10,col=brewer.pal(8,"Dark2"),rot.per=0.3 )
#图 3-9-3(a)
#wordcloud(row.names(df) , df$Paper2 , min.freq=10,col=brewer.pal(8,"Dark2"),rot.per=0.3 )
#图 3-9-3(b

错误是：> commonality.cloud(term.matrix,max.words=100,random.order=FALSE,color="#E7298A")
Error in is_overlap(x1, y1, sw1, sh1, boxes) :
程序包'Rcpp_precious_remove'不提供'Rcpp'这样的函数

[chenwenqi228]

我也是频繁报错，后来结合chatgpt反复修改，发现问题主要出在：
1）Rstudio版本和命令版本不兼容，需要检查，改更新的就更新；
2）paper1和paper2不是UTF-8格式，需要设定一下
3）过滤停止词库的命令有bug，需要补充：
corp <- tm_map(corp, content_transformer(function(x) iconv(enc2utf8(x), sub = "byte")))
corp <- tm_map(corp, function(x){removeWords(x,stopwords("en"))})

现在这个版本的可以运行：
library(tm)
library(wordcloud)
Paper1 <- paste(scan("R语言可视化之美/第3章_类别比较型图表/Paper1.txt", what = character(0), sep = "", encoding = "UTF-8"), collapse = " ")
Paper2 <- paste(scan("R语言可视化之美/第3章_类别比较型图表/Paper2.txt", what = character(0), sep = "", encoding = "UTF-8"), collapse = " ")
#补充encoding = "UTF-8"，防止文件识别格式错误

tmpText<- data.frame(c(Paper1, Paper2),row.names=c("Text1","Text2"))
df_title <- data.frame(doc_id=row.names(tmpText),
text=tmpText$c.Paper1..Paper2.)
ds <- DataframeSource(df_title)
#创建一个数据框格式的数据源，首列是文档id(doc_id),第二列是文档内容
corp <- VCorpus(ds)
#加载文档集中的文本并生成语料库文件
corp<- tm_map(corp,removePunctuation) #清除语料库内的标点符号
corp <- tm_map(corp,PlainTextDocument) #转换为纯文本
corp <- tm_map(corp,removeNumbers) #清除数字符号

corp <- tm_map(corp, content_transformer(function(x) iconv(enc2utf8(x), sub = "byte")))
corp <- tm_map(corp, function(x){removeWords(x,stopwords("en"))}) #过滤停止词库
term.matrix <- TermDocumentMatrix(corp)
#利用TermDocumentMatrix()函数将处理后的语料库进行断字处理，生成词频权重矩阵

term.matrix <- as.matrix(term.matrix) #频率
colnames(term.matrix) <- c("Paper1","paper2")
df<-data.frame(term.matrix)
write.csv(df,'term_matrix.csv') #导出两篇文章的频率分析结果

#---------------------------------------导入数据------------------------------------------
df<-read.csv('term_matrix.csv',header=TRUE,row.names=1)

#----------------------------------------两篇文章数据的对比-------------------------------------------------------------
comparison.cloud(df, max.words=300, random.order=FALSE, rot.per=.15, c(4,0.4), title.size=1.4)

comparison.cloud(df,max.words=300,random.order=FALSE,colors=c("#00B2FF", "red"))

commonality.cloud(df,max.words=100,random.order=FALSE,color="#E7298A")

comparison cloud

comparison.cloud(df, random.order=FALSE,
colors = c("#00B2FF", "red", "#FF0099", "#6600CC"),
title.size=1.5, max.words=500)

#-------------------------------------单篇文章数据的展示-----------------------------------------------------------------
#Colors<-colorRampPalette(rev(brewer.pal(9,'RdBu')))(length(df$Paper1>10))
wordcloud(row.names(df) , df$Paper1 , min.freq=10,col=brewer.pal(8, "Dark2"), rot.per=0.3 )