Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Hash collisions with deep searches

Author: Robert Hyatt

Date: 15:13:08 12/19/02

Go up one level in this thread


On December 19, 2002 at 17:36:22, Joshua Haglund wrote:

>On December 19, 2002 at 17:31:36, Robert Hyatt wrote:
>
>>On December 19, 2002 at 17:22:15, Joshua Haglund wrote:
>>
>>>On December 19, 2002 at 16:50:32, Dann Corbit wrote:
>>>
>>>>Suppose that we are searching at long time control, and the hash table rapidly
>>>>fills up.  This can throw move ordering in the toilet, if we can't find
>>>>transpositions any more.
>>>>
>>>>What sort of replacement scheme is best for very long searches (several
>>>>minutes)?
>>>
>>>I like to take stabs at topics even if I'm unaware of the topic.
>>>
>>>The first way I can think of dealing with this is to clear the hash table and
>>>continue with the search.
>>
>>Think about that for a second.  It would actually make things _worse_ not
>>better, because without clearing it, you at least have some useful info.
>
>How about clearing the first half of the hash table when it gets to max?
>>

Same problem.  The information in that is important.  Dumping it is just as
bad as not storing anything more...



>>>
>>>Second, how about making the hash table bigger.
>>>
>>>Third, create a second hash table.
>>>if (hash table = full) {
>>>switch to table2}
>>>else {
>>>};return search
>>>
>>>I hope this atleast humors you,
>>
>>The problem is that the size of the table is probably already at max...
>>
>>
>>>
>>>Joshua
>>>toneewa@yahoo.com



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.