Why are there so many random generators in R?

Question

A quick look at the help page of the random number generator of R shows that you can choose among 7 predefined generators (Wichmann-Hill, Marsaglia-Multicarry, Super-Duper, Mersenne-Twister, Knuth-TAOCP-2002, Knuth-TAOCP, L'Ecuyer-CMRG).

?Random

The default is Mersenne-Twister, which seems to be very good. So why would you ever need to use another one?

Answer 1

One case where this RNG cornucopia is handy is when you're rewriting or comparing software that rely on different RNGs. One example might be porting R code to C++. You want to pin down why you're getting different results and it helps to hold as many things constant as possible.