R Aborting Only used 200MB of Memory

[nerdybaseballguy]

Hi all. RStudio is constantly aborting whenever I run a custom function that manipulates a dataframe and replaces it with another dataframe. (i.e. Dataframe1=CustomFunction(Dataframe1) but RStudio aborts whenever I run the function. When I run mem_used it shows I am only using 200MB of memory, far below the limit. If anyone could please shed some light on the situation, I would appreciate it.

Thanks

[nirgrahamuk]

Use debug() or browser() to identify the code statement that causes the abort.

[nerdybaseballguy]

There isn't a problem with the function itself. The function runs fine. The assignment to the dataframe is where R aborts, but other times it works fine. I just do not understand why R is aborting. I would usually think I was running out of RAM, but I'm only using 200MB.

[nirgrahamuk]

@nerdybaseballguy

Dataframe1=CustomFunction(Dataframe1)

Let's be clear, are you saying that the above always aborts. And

CustomFunction(Dataframe1)

Never aborts? and prints output to the console?

[nerdybaseballguy]

Yes, that is correct. Very strange

[nirgrahamuk]

Unfortunately I think that without access to the function code it will be impossible for anyone to debug/diagnose for you.

[nerdybaseballguy]

Here is the function. There are other custom functions that this function calls, but all the dependencies run 100% good all the time. Here is the function. I apologize for the length.

ShortSim=function(SimmedSchedule,Year,runs){
rm(FinalR)
Schedule=SimmedSchedule
rm(SimmedSchedule)
NotPlayed=Schedule %>%
  filter(Simmed==1)
Schedule=Schedule %>%
  filter(Simmed==0)

NotPlayed$VisitingTeam=str_trim(NotPlayed$VisitingTeam,side = "both")
NotPlayed$HomeTeam=str_trim(NotPlayed$HomeTeam,side = "both")

for(i in 1:nrow(NotPlayed)){
  tryCatch({
    message("Simulating ",paste(NotPlayed[i,3]," @ ",NotPlayed[i,2]," on ",NotPlayed[i,1],"   *** ",round((i/nrow(NotPlayed))*100,digits = 1),"% Complete ***",sep = ""))
    hold=SimGame(NotPlayed[i,3],NotPlayed[i,2],runs)
    #Check to make sure aggregate results from sims to not come out equal which would result in a tie. If they do, chose randomly
    if(hold[[1]] == hold[[2]]){
      if(runif(1)>=.5){
        hold[[1]]=1
        hold[[2]]=0
      }else{
        hold[[1]]=0
        hold[[2]]=1
      }
    }
    NotPlayed[i,5]=hold[[1]]
    NotPlayed[i,4]=hold[[2]]
  },error=function(e){})
}
beep(2)

NotPlayed=NotPlayed[complete.cases(NotPlayed),]
NotPlayed=NotPlayed %>%
  select("date","HomeTeam","VisitingTeam","HomeScore","VisitingScore","Neutral")
NotPlayed$Simmed=1
Schedule=Schedule %>%
  select("date","HomeTeam","VisitingTeam","HomeScore","VisitingScore","Neutral")
Schedule$Simmed=0
Schedule=Schedule[complete.cases(Schedule),]
SimmedSchedule=rbind.data.frame(Schedule,NotPlayed)
rm(NotPlayed)
SimmedSchedule=SimmedSchedule %>%
  arrange(date)
beep(1)
#FinalR=ApplyModels(CalculateConfChamp(CalculateRPI(SimmedSchedule)))
SimmedSchedule=CalculateRPI(SimmedSchedule)
#write.csv(Result,"F:\\Result.csv")
rm(SimmedSchedule)
gc()
beep(2)
#Result=read.csv("F:\\Result.csv")
SimmedSchedule=CalculateConfChamp(SimmedSchedule)
#rm(Result)
FinalR=ApplyModels(SimmedSchedule)
#rm(ConfR)

#rm(ConfR)
#rm(Result)
#rm(Schedule)
beep(2)
 return(FinalR) 
}

[nerdybaseballguy]

I've tried various gimmicks such as writing it to a CSV, bringing it back and removing various things throughout the function. That's why there are so many things commented out.