R Notebook

library(tm)
library(dplyr)
library(tidyr)
quijote<-file("quijote.txt","r",encoding = "UTF-8")
quijote_lines<-readLines(quijote)
#############Parte 1################
#quijote_words<-paste(quijote_lines,collapse = " ")
#quijote_words<-strsplit(quijote_words,split = " ")
#length(quijote_words[[1]]) #num pal
#length(unique(quijote_words[[1]])) #num pal unique
#quijote_words<-sapply(quijote_words[[1]],"removePunctuation",USE.NAMES = FALSE) 
#quijote_words<-sapply(quijote_words,"tolower",USE.NAMES = FALSE)
#quijote_words<-sapply(quijote_words,"stripWhitespace",USE.NAMES = FALSE)
#quijote_words<-sapply(quijote_words,"removeNumbers",USE.NAMES = FALSE)
#quijote_words<-quijote_words[quijote_words!=""]
  
#table(quijote_words) %>% as.data.frame() %>% View()
        
#############Parte 2################
quijote_src<-VectorSource(quijote_lines)
quijote_corpus<-VCorpus(quijote_src)
clean_corpus<-function(corpus){
  corpus<-tm_map(corpus,stripWhitespace)
  corpus<-tm_map(corpus,removePunctuation)
  corpus<-tm_map(corpus,content_transformer(tolower))
  corpus<-tm_map(corpus,removeNumbers)
  return(corpus)
}
quijote_corpus<-clean_corpus(quijote_corpus)
#quijote_dtm<-DocumentTermMatrix(quijote_corpus)
#quijote_dtm_m<-as.matrix(quijote_dtm)
#quijote_tdm<-TermDocumentMatrix(quijote_corpus)
#quijote_tdm_m<-as.matrix(quijote_tdm)
BigramTokenizer<-function (x){
  unlist(lapply(ngrams(words(x),2),paste,collapse = " "),use.names = FALSE)
}
tdm<-TermDocumentMatrix(quijote_corpus,control = list(tokenize = BigramTokenizer))

closing unused connection 4 (quijote.txt)

#m<-as.matrix(tdm)
#rowSums(as.matrix(tdm)) %>% as.data.frame() %>% View
#bigram<-as.data.frame(bigram)
#names(bigram)<-c("frec","bigram")
#bigram<-as.data.frame((bigram))
#bigram$bigram<-as.character((bigram$bigram))
#bigram<-separate(bigram,bigram,into = c("word1", "word2"))
#sentence<- function(palabra = "venta", n = 10){
 # oracion <- palabra
  #for(i in 1:n){
   # NextWord<-bigram %>% 
    #filter(word1 == palabra) %>% 
    #arrange(desc(freq)) %>% head(1) %>% select(word2) %>% as.character()
  #paste(c(oracion,NextWord),collapse = " ")
  #palabra<-NextWord
  #}
#}
sentence_maxpro<- function(palabra = "venta", n = 10){
  oracion <- palabra
  for(i in 1:n){
    NextWord<-bigram %>% 
    filter(word1 == palabra) %>% 
    arrange(desc(freq)) %>% head(5) 
    NextWord<-sample(NextWord$word2,size = 1, prob = NextWord$freq/sum(NextWord$freq))
    
    
    
  paste(c(oracion,NextWord),collapse = " ")
  palabra<-NextWord
  }
}