Fastest way to cross-tabulate two massive logical vectors in R

Question

For two logical vectors, x and y, of length > 1E8, what is the fastest way to calculate the 2x2 cross tabulations?

I suspect the answer is to w

Answer 1

Here is an answer using Rcpp sugar.

N <- 1e8
x <- sample(c(T,F),N,replace=T)
y <- sample(c(T,F),N,replace=T)

func_logical  <- function(v1,v2){
    return(c(sum(v1 & v2), sum(v1 & !v2), sum(!v1 & v2), sum(!v1 & !v2)))
}


library(Rcpp)
library(inline)

doCrossTab1 <- cxxfunction(signature(x="integer", y = "integer"), body='
  Rcpp::LogicalVector Vx(x);
  Rcpp::LogicalVector Vy(y);
  Rcpp::IntegerVector V(4);

  V[0] = sum(Vx*Vy);
  V[1] = sum(Vx*!Vy);
  V[2] = sum(!Vx*Vy);
  V[3] = sum(!Vx*!Vy);
  return( wrap(V));
  '
, plugin="Rcpp")

system.time(doCrossTab1(x,y))

require(bit)
system.time(
{
xb <- as.bit(x)
yb <- as.bit(y)
func_logical(xb,yb)
})

which results in:

> system.time(doCrossTab1(x,y))
   user  system elapsed 
  1.067   0.002   1.069 
> system.time(
+ {
+ xb <- as.bit(x)
+ yb <- as.bit(y)
+ func_logical(xb,yb)
+ })
   user  system elapsed 
  1.451   0.001   1.453 

So, we can get a little speed up over the bit package, though I'm surprised at how competitive the times are.

Update: In honor of Iterator, here is a Rcpp iterator solution:

doCrossTab2 <- cxxfunction(signature(x="integer", y = "integer"), body='
  Rcpp::LogicalVector Vx(x);
  Rcpp::LogicalVector Vy(y);
  Rcpp::IntegerVector V(4);
    V[0]=V[1]=V[2]=V[3]=0;
  LogicalVector::iterator itx = Vx.begin();
  LogicalVector::iterator ity = Vy.begin();
  while(itx!=Vx.end()){
    V[0] += (*itx)*(*ity);
    V[1] += (*itx)*(!*ity);
    V[2] += (!*itx)*(*ity);
    V[3] += (!*itx)*(!*ity);    
    itx++;
    ity++;
  }
  return( wrap(V));
  '
, plugin="Rcpp")

system.time(doCrossTab2(x,y))
#   user  system elapsed 
#  0.780   0.001   0.782