R - Statistical Analysis

#Load & Assign Files
TrainDataOrange <- read.arff("D:/Downloads/orange_train.arff")
colnames(TrainDataOrange)[colnames(TrainDataOrange)=="Var230"] <- "class"
TestDataOrange <- read.arff("D:/Downloads/orange_test.arff")
colnames(TestDataOrange)[colnames(TestDataOrange)=="Var230"] <- "class"

#Obtain Class Values
Actual <- TestDataOrange[, 230]

#Ranking
A <- GainRatioAttributeEval(class ~ . , data = TrainDataOrange,na.action=NULL )
ranked_list <- A[order(A)]
A[order(-A)]
HighestChurnScore <- 0

#Feature Drop
for (i in seq(225,5,-5)) {
  D <- (225-i)
  s <- ranked_list[1:D]
  cols.dont.want <- c(names(s))
  TrainDataOrange1 <- TrainDataOrange[, !names(TrainDataOrange) %in% cols.dont.want, drop=T]
  
  # Classifier
  NB <- make_Weka_classifier("weka/classifiers/bayes/NaiveBayes")
  classifier <- NB(class~.,data=TrainDataOrange1,na.action=NULL)
  TestDataOrange1 <- TestDataOrange[, !names(TrainDataOrange) %in% cols.dont.want, drop=T]
  
  pred<-predict(classifier,TestDataOrange1, na.action=NULL,seed=1)
  
  #Scoring
  P11<-0
  P12<-0
  P21<-0
  P22<-0
  Actual[]
  for(K in seq(1,1667))
  {
    if(Actual[K]==-1) #LOYAL
    {
      if(pred[K]==-1)
      {
        P11<-P11+1
      }
      else
      {
        P12<-P12+1
      }
    }
    else if(Actual[K]==1) #CHURN
    {
      if(pred[K]==1)
      {
        P22<-P22+1
      }
      else
      {
        P21<-P21+1
      }
    }
  }
  
  #F_Score
  Prec_1<-(P11/(P11+P21))
  Prec_2<-(P22/(P22+P12))
  Recall_1<-(P11/(P11+P12))
  Recall_2<-(P22/(P22+P21))
  
  
  F_1<-(5*Prec_1*Recall_1)/(4*Prec_1+Recall_1)
  F_2<-(5*Prec_2*Recall_2)/(4*Prec_2+Recall_2)
  
  if (F_1 >= 0.95) {
    if (F_2 > HighestChurnScore) {
      HighestChurnScore <- F_2
      cat("Number of features selected:",i,"\n")
      cat("This is the F2 Loyal score",F_1,"\n")
      cat("This is the F2 Churn score",F_2,"\n")
    }
    cat("…\n")
  }
}