R Studio crashes after using Rcpp function

Question

What I'm trying to do is find matching dates between multiple large matrices. And what I want my C++ code to return is row indices where matches are found

I am completely new to C++ and I have found it extremely useful to speed up my R code.

My code seems to work in R Studio but crashes after the Rcpp function has been used through some iterations in another loop in my R code

Here is some example data

baseflow_mat[[1]] is a matrix formatted as so

baseflow_mat[[2]] is an example of where I want to find matches

baseflow_mat<-list()

baseflow_mat[[1]]<-data.frame(year=c(1992,1992,1992,1992),month=c(7,7,7,7),day=c(5,10,13,17),flow=c(50,60,59,33))

baseflow_mat[[2]]<-data.frame(year=c(1992,1992,1992,1992,1992,1992,1992,1992),month=c(7,7,7,7,7,7,8,8),day=c(4,10,13,18,26,27,2,6),flow=c(50,60,59,33,45,40,55,52))

And I want to find matching dates across all 170 large matrices of baseflow_mat

baseflow_mat[[2]] is an example of where I want to find matches

So what I want my C++ code to return is the row indices of baseflow_mat_2 of all matches from baseflow_mat_1. This works fine with small matrices but once I start to use all my data it starts to crash. My actual data has anywhere from 500 to 3000 rows in each matrix and I want to find matches for 170 separate matrices so 170*170 about 28900 results

Here is my R code

library(Rcpp)
sourceCpp("Source1.cpp")

big_match<-list()
for(i in seq(1,2)){#length(baseflow_mat))){
  match_baseflow_list<-list()
  for(j in seq(1,2)){#length(baseflow_mat))){
     matches_wzeros<-matchRows(nrow(baseflow_mat[[j]][,1:3]),nrow(baseflow_mat[[i]][,1:3]),baseflow_mat[[j]][,1:3],baseflow_mat[[i]][,1:3])
     matches<-matches_wzeros[matches_wzeros>0]
     match_baseflow_list[[j]]<-baseflow_mat[[j]][matches,]
   }
  big_match[[i]]<-match_baseflow_list
}

Here is my C++ code

// [[Rcpp::export]]

NumericVector matchRows(int rowSize, int matchRowSize, DataFrame nonMatchDF, DataFrame matchDF)
{
//0 is for year, 1 is for month, 2 is for day for both DF
Rcpp::NumericVector nonmatchYear = nonMatchDF[0];
Rcpp::NumericVector nonmatchMonth = nonMatchDF[1];
Rcpp::NumericVector nonmatchDay = nonMatchDF[2];

Rcpp::NumericVector matchYear = matchDF[0];
Rcpp::NumericVector matchMonth = matchDF[1];
Rcpp::NumericVector matchDay = matchDF[2];

Rcpp::NumericVector indexMatrix(matchRowSize*rowSize);


//j is for going through the nonmatch dataframe
int j;

//i is for going through the other DF
int i;

//addIndex is to add correctly to the vector
int addIndex = 0;

for (i = 0; i < matchRowSize; i++)
{
    //Rcpp::NumericVector indexMatch(rowSize);


    for (j = 0; j < rowSize; j++)
    {

        if ((matchYear[i] == nonmatchYear[j]) && (matchMonth[i] == nonmatchMonth[j]) && (matchDay[i] == nonmatchDay[j]))
        {
            indexMatrix[addIndex + j] = j + 1 ;
            //indexMatrix(i, j) = j+1;
        }
    }

    addIndex = addIndex + (j+1);
}

return indexMatrix;
  }

And sessionInfo() output

R version 3.1.1 (2014-07-10)
Platform: x86_64-w64-mingw32/x64 (64-bit)

locale:
[1] LC_COLLATE=English_United States.1252 
[2] LC_CTYPE=English_United States.1252   
[3] LC_MONETARY=English_United States.1252
[4] LC_NUMERIC=C                          
[5] LC_TIME=English_United States.1252    

attached base packages:
[1] stats     graphics  grDevices utils    
[5] datasets  methods   base     

other attached packages:
[1] Rcpp_0.11.6

loaded via a namespace (and not attached):
[1] tools_3.1.1

Answer 1

What you are doing inside the innermost loop is essentially a merge. If you need it to be faster, use dplyr::left_join (a fast alternative for merge).

big_match<-list()
for(i in seq(1,2)){#length(baseflow_mat))){
  match_baseflow_list<-list()
  for(j in seq(1,2)){#length(baseflow_mat))){
    match_baseflow_list[[j]] <- merge(baseflow_mat[[1]], baseflow_mat[[2]], by=c("year", "month", "day"))
  }
  big_match[[i]]<-match_baseflow_list
}