0

After I prepare my data base as shown in the code below, I tried to split the data into training and testing. So I uesd createDataPartition function to do this.

My R scripts in c#: 

// the data base
DataFrame BASE = engine.Evaluate("BASE<-data.frame(Type_peau,PEAU_CORPS,SENSIBILITE,IMPERFECTIONS,BRILLANCE,GRAIN_PEAU,RIDES_VISAGE,ALLERGIES,MAINS,INTERET_ALIM_NATURELLE,INTERET_ORIGINE_GEO,INTERET_VACANCES,INTERET_COMPOSITION,INT_AGE,priorite2,priorite1,MILRES,Profil_Select,Age,Nbre_gift,Achat_client)").AsDataFrame();

// Converting categorical variables into quantitatives

engine.Evaluate("must_convert<-sapply(BASE,is.factor)").AsExpression();
engine.Evaluate("M2<-sapply(BASE[,must_convert],unclass)").AsExpression();
engine.Evaluate("out<-cbind(BASE[,!must_convert],M2)").AsExpression();
engine.Evaluate("out=unlist(lapply(out,as.numeric))").AsExpression();
engine.Evaluate("out=as.data.frame(out)").AsExpression();
engine.Evaluate("out$Achat_client=sapply(out$Achat_client,function(x) (if (x==1) {'Fidèle'} else {'Non Fidèle'}))").AsExpression();
engine.Evaluate("out$Achat_client=as.factor(unlist(out$Achat_client))").AsExpression();

// Creating new data base

DataFrame M1 =engine.Evaluate("M1=out").AsDataFrame();

// Creating the index of splitting

engine.Evaluate("val_index<-createDataPartition(M1$Achat_client,p=.75,list=FALSE)").AsExpression();

I got just an error in the last line of the code above :

Error in createDataPartition(M1$Achat_client, p = 0.75, list = FALSE) : y must have at least 2 data points

I searched and I find this logical solution here r- caret package error- createDataParition no observation.

So I remplaced M1$Achat_client by M1[[21]] and also by M1[,21] in other time just to try if it will work, But the error still shown !

Other thing, perphaps my origin script work very well in R in way of using M1$Achat_client . I'm surprised why in c# i have these kinds of error? Please how do you explain this?

Craig W.
  • 17,838
  • 6
  • 49
  • 82
User2018
  • 61
  • 10

1 Answers1

0

Well, As a solution I remplace the use of the function createDataPartition to split the data base by using other method proposed here by dickoa

How to split data into training/testing sets using sample function

. It has the same purpose! and it works very well in my case of use!

User2018
  • 61
  • 10