sim_crossovers.cpp

#include <Rcpp.h>
using namespace Rcpp;

// [[Rcpp::export]]
NumericVector sim_crossovers_orig(const double L)
{
    int n_xo;

    n_xo = R::rpois(L/100.0);

    NumericVector tmp = runif(n_xo, 0.0, L);
    return tmp.sort();
}

// [[Rcpp::export]]
NumericVector sim_crossovers_new(const double L)
{
    int n_xo;

    n_xo = R::rpois(L/100.0);

    NumericVector tmp(0);
    if(n_xo > 0) tmp = runif(n_xo, 0.0, L);
    return tmp.sort();
}

Code to explore a change in Rcpp from version 0.12.15 to 0.12.16.

With Rcpp ver 0.12.15, these two things give the same results.

Rcpp::sourceCpp("sim_crossovers.cpp")

set.seed(20180318)
z <- replicate(10000, sim_crossovers_orig(100))
table(sapply(z, length))
##    0    1    2    3    4    5    6
## 3661 3681 1837  631  151   35    4

set.seed(20180318)
z <- replicate(10000, sim_crossovers_new(100))
table(sapply(z, length))
##    0    1    2    3    4    5    6
## 3661 3681 1837  631  151   35    4

But with Rcpp ver 0.12.16, the former gives a ton of 0-length vectors, while the latter works fine, though it gives somewhat different results from ver 0.12.15. (The tables being printed should be 10,000 draws from a Poisson distribution with mean 1.)

Rcpp::sourceCpp("sim_crossovers.cpp")

set.seed(20180318)
z <- replicate(10000, sim_crossovers_orig(100))
table(sapply(z, length))
##    0    1    3
## 9994    5    1

set.seed(20180318)
z <- replicate(10000, sim_crossovers_new(100))
table(sapply(z, length))
##    0    1    2    3    4    5    6
## 3637 3698 1846  621  152   41    5