builldsp

buildSp.R

library(sqlrutils) # Contains the utiity that lets us create a stored procedure automatically from R code 

buildCustomerTree <- function(trainingDs) {  
  
  #
  # Make factors out of these numbers otherwise they get treated as continuos values 
  #
  trainingDs$TotalChildren <- as.factor(trainingDs$TotalChildren)
  trainingDs$NumberChildrenAtHome <- as.factor(trainingDs$NumberChildrenAtHome)
  trainingDs$NumberCarsOwned <- as.factor(trainingDs$NumberCarsOwned)
  trainingDs$YearsSinceFirstPurchase <- as.factor(trainingDs$YearsSinceFirstPurchase)
  
  trainingDs$AgeRanges <- cut(
                              trainingDs$Age,breaks=c(0,30,50,70,90),
                              labels=c("11-30","31-50","51-70","71+")
                              )
  trainingDs$IncomeCategories <- cut(
                                      x=trainingDs$YearlyIncome,
                                      breaks=c(0,20000,50000,70000,100000,250000),
                                      labels=c("Low","Lower","Middle","Upper","Wealthy")
                                      )

  tr <- rxDTree(
      IsCardUser~
        
        Gender+
        Occupation+
        MaritalStatus+
        IsHomeOwner+
        TotalChildren+
        NumberCarsOwned+
        AgeRanges+
        CommuteDistance+
        IncomeCategories+
        CountryRegionCode+
        Education,
      
      data=trainingDs,
      method="class",
      overwrite = TRUE,
      reportProgress = 0,
      cp = 0.03
    )
  
    trained_model <- rxSerializeModel(tr)
    return(list(trained_model=trained_model))
}