Bernd wroteI am cheating too. Instead of a hash table in the commonly known sense my "hash table" is implemented as a trie.
So don't be surprised when your program is rejected :-)
Mailing lists / fpc-other / 2011-05
k-nucleotide implementation (was: Which programming language is fastest?)
2 messages · 2 people · –
- Isaac Gouy 1
- Jeppe Johansen 1
Reply by email Subscribe to this list Your subscription
Read-only archive. Use j and k to move between messages.
Isaac Gouy wrote 9 lines
On 08 May 2011, Bernd wrote:
> I am cheating too. Instead of a hash table in the
> commonly known sense my "hash table" is implemented as a trie.
So don't be surprised when your program is rejected :-)
_______________________________________________
fpc-other maillist -
http://lists.freepascal.org/mailman/listinfo/fpc-other
Just a note to the original post about the k-nucleotide implementation:
I've been writing on a crude sampling profiler lately which indicates(on a dual core x86) that 22% of the time is spent in ht_hashcode, 25% is spent in the loop in FindNew, and 28 is spent in the RTL(looks like CompareByte)
Those measurements are far from scientific, but I think it gives a hint that maybe the algorithm is wrong(and maybe CompareChar), and not so much the code generation/optimization