in reply to Re^3: Efficient Unique Nested Combinations
in thread Efficient Unique Nested Combinations

If I understand you — and I very well may not — you'd be generating the whole set and then letting M::C filter it down. Can you illustrate what you're talking about, in Perl?

Caution: Contents may have been coded under pressure.
  • Comment on Re^4: Efficient Unique Nested Combinations

Replies are listed 'Best First'.
Re^5: Efficient Unique Nested Combinations
by Moron (Curate) on Jun 28, 2007 at 11:29 UTC
    No I meant to just iterate nested next-combinations but it comes to my attention that actually both loops were degenerate (nCn and nC1) so that M::C would be doing nothing more than returning next elements in arrays, which might as well be handrolled anyway. To assure unique selections from inner arrays, the OPs hash idea seems now as good as any.