Re: Array vs. Hash for sparsely integer-indexed data

[flexvault]

puterboy,

In particular, I am asking: 1. What (if any) storage inefficiency (relative to hashes) is created in perl when using arrays for sparsely indexed data? 2. What (if any) lookup inefficiency (relative to arrays) is created in perl when using hashes for positive integer indexed keys?

I'm sure you'll get more technical answers, but I use hashes now even when the indexing is sequential. The ability to use 'defined' or 'exists' to see if the key is present greatly improves the speed of lookups. And think of the storage benefits, since if you do the following:

   my @array; $array[100] = 246;
   my %hash;  $hash{"100"} = 246;

The array will use 101 ( 0..100 ) locations and the hash will allocate 8 (hashes are allocated in powers of 2).

But why not 'benchmark' it for yourself for your exact situation and know for sure :-)

Good Luck...Ed