Share via

Error 1000 on Microsoft Machine Learning

Tehila 1 Reputation point
2022-01-28T09:20:46.673+00:00

Hi everyone,
I'm using Azure Studio (Microsoft Machine Learning Studio (classic)) to run R script.

I'm trying to run a loop to find auc on 1,000 times.

The loop and the code run perfectly when the iterator is up to 100, and fail on 1,000.

The error is:
169422-image.png

Thanks for your help!

The full script is:
install.packages ("AUC", repos= "https://cran.microsoft.com/snapshot/2022-01-13/")

# Map 1-based optional input ports to variables  
universal_bank.df <- maml.mapInputPort(1) # class: data.frame  
  
ComputeUAC <- function( seed ){  
  #split data  
  set.seed( seed )  
  train.index <- sample(1:dim( universal_bank.df )[1], dim( universal_bank.df )*0.6)  
  train.df <- universal_bank.df[train.index , ]  
  valid.df <- universal_bank.df[-train.index , ]  
    
  library(party)  
  tr <- ctree( Personal.Loan ~. , data = train.df  )  
  plot(tr, type = "simple")  
  
  pred <- predict( tr, newdata = valid.df )  
    
  
  library(AUC)  
  r <- roc( pred, as.factor(valid.df$Personal.Loan) )  
  #plot(r)  
  UAC <- auc(r)  
    
  return ( UAC )  
  
}  
  
UAC_arr <- c()  
for( seed in c(1:10000)){  
  UAC_arr <- c( UAC_arr , ComputeUAC( seed ) )  
}  
  
boxplot(UAC_arr);  
  
UAC_arr.df <- as.data.frame( UAC_arr )  
  
# Select data.frame to be sent to the output Dataset port  
maml.mapOutputPort("UAC_arr.df");  
Azure Machine Learning

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.