Message5882

Author jendrik
Recipients andrew.coles, erez, florian, jendrik, malte, silvia
Date 2016-12-14.15:47:41
Content
Yes, the experiment only compares the old hash function and hash table to the new 
hash function and hash table. On the instance I tested locally, the old hash 
function basically produced ascending hash values, which is very bad for 
hopscotch hashing. 

I have split off the issues as you suggested:

A) issue692: operator IDs instead of generating operator pointers
B) issue693: new hash function
C) issue694: new hash table implementation
D) issue695: space improvement for LM-Count
History
Date User Action Args
2016-12-14 15:47:41jendriksetmessageid: <1481726861.12.0.198725132258.issue213@unibas.ch>
2016-12-14 15:47:41jendriksetrecipients: + jendrik, malte, erez, andrew.coles, silvia, florian
2016-12-14 15:47:41jendriklinkissue213 messages
2016-12-14 15:47:41jendrikcreate