On 13/10/2010 12:49 PM, scott wrote:
>> One immediate and obvious problem is that the buckets might fill very
>> unequally, depending on the distribution of the data to be sorted.
>
> Is this really a problem? It shouldn't affect the run-time of the radix
> sort algorithm at all.

Hmm, yes. The running time of a radix sort is proportional to the length 
of the keys, and regardless of the volume of data. Which means that even 
sorting a small volume of data could be quite slow if the keys are large.

It also doesn't help that the keys I want to sort by are floating-point 
numbers... although I suppose you could truncate them to integers and do 
it that way.

Post a reply to this message