Mise au format de type LETOR

gistfile1.r

###############################################################################
####### MISE AU FORMAT LETOR ##################################################
ST4<-NULL
ST2<-NULL

#ec2-79-125-53-143.eu-west-1.compute.amazonaws.com

library(dummies)

library(gtools)
letor <- defmacro(colonne, table, expr={
  #table[,colonne]<-paste(grep(colonne, colnames(table)),":",table[,colonne])
  table[,colonne]<-paste(match(colonne,names(table)),":",table[,colonne])
  table[,colonne] <- gsub(pattern = " ", replacement = "",  x = table[,colonne])
  table[,colonne] <- gsub(pattern = "\\d\\d.NA$", replacement = "",  x = table[,colonne])
  table[,colonne] <- gsub(pattern = "\\d.NA$", replacement = "",  x = table[,colonne])
})


t<-as.data.frame(dummy(ST3$site_id))

t1<-as.data.frame(dummy(ST3$visitor_location_country_id))

t2<-as.data.frame(dummy(ST3$prop_country_id))

t3<-as.data.frame(dummy(ST3$prop_starrating))

t4<-as.data.frame(dummy(ST3$prop_review_score))

#t5<-as.data.frame(dummy(ST3$srch_destination_id))


ST3$site_id <- NULL
ST3$visitor_location_country_id <- NULL
ST3$prop_country_id <- NULL
ST3$prop_starrating <- NULL
ST3$prop_review_score <- NULL
ST3$srch_destination_id <- NULL
ST3$V4 <- NULL
ST3$prop_id <- NULL

ST4<-cbind(ST3,t,t1,t2,t3,t4)

ST4$srch_id <- paste("qid",":",ST4$srch_id)
ST4$srch_id <- gsub(pattern = " ", replacement = "",  x = ST4$srch_id)

ST3 <-NULL
t<-NULL
t1<-NULL
t2<-NULL
t3<-NULL
t4<-NULL
t5<-NULL

for (i in 3:302) {
  
  letor(colnames(ST4)[i],ST4)
  print(i)
}


write.table(ST4, "/home/alfard/train.dat", sep="\t",quote=FALSE,col.names=FALSE,row.names=FALSE)