[R] textmining-Clusting-dendogram : invalid dendrogram input

[vegas55]

HI All, 

Please I need help with the following. I use the TM package on text mining
purpuse. Everything works fine until the stage of trying to do a dendogram.
R gives this message (See the end of the script) :

Error in graphics:::plotHclust(n, merge, height, order(x$order), hang,  : 
  invalid dendrogram input


# Lire le fichier texte
filePath <- "C:\\BI\\projet supervise\\Sondage satisfaction
PV//PVTextMining.txt"
text <- readLines(filePath)

# Charger les données comme un corpus
docs <- Corpus(VectorSource(text))
inspect(docs)

# Convertir le texte en minuscule
docs =tm_map(docs,tolower)
#docs <- tm_map(docs, content_transformer(tolower))

# Supprimer les nombres
docs <- tm_map(docs, removeNumbers)

# Supprimer les mots vides français
docs <- tm_map(docs, removeWords, stopwords("french"))

# Supprimer votre propre liste de mots non désirés
docs <- tm_map(docs, removeWords, c("non") 

# Supprimer les ponctuations
docs <- tm_map(docs, removePunctuation)

# Supprimer les espaces vides supplémentaires
docs <- tm_map(docs, stripWhitespace)

# Text stemming
docs <- tm_map(docs, stemDocument)

#Enlève des caractères spéciaux

docs = tm_map(docs, function(x) gsub("\\W", " ", x))

#Etape 4: Construire la matrice des mots

dtm <- TermDocumentMatrix(docs, control=list(wordLenghts=c(1, Inf)))

##############cluster

#######methode 1

#Hierarchical Clustering

dtm2 <- removeSparseTerms(dtm, sparse=0.95)
matrix2 <- as.matrix(dtm2)

#Now we compute the distance matrix for the hclust() function.

distMatrix <- dist(scale(matrix2))

cluster <- hclust(distMatrix, method="ward")
pp

pp <- dist(matrix2, method="ward")

#plot dendogram euclidean
windows()

plot(cluster, hang=-1, main="Clusters")

Error in graphics:::plotHclust(n, merge, height, order(x$order), hang,  : 
  invalid dendrogram input

Thanks a lot for your help ! 

YR






--
View this message in context: http://r.789695.n4.nabble.com/textmining-Clusting-dendogram-invalid-dendrogram-input-tp4712145.html
Sent from the R help mailing list archive at Nabble.com.