2008-04-20
| 08:07 | rhickey | Modius: in biasing for insertion/removal is it not going to be persistent? |
| 10:34 | jteo | projecteuler is good fun for learning clojure |
| 12:41 | Modius | rhickey: I probably didn't explain that well. Structurally the same - I meant the functions take lists of removals/additions to be done en masse - it's still functional, and still structurally the same. |
| 13:01 | rhickey | Modius: are the bulk adds/removes for performance or convenience? |
| 13:01 | Modius | rhickey: I may not have tested this as much as you have - but they are in the vain attempt to cycle out less buckets in bulk-insert cases (performance) - for convenience itself, either can be built out of the other. |
| 13:03 | rhickey | not sure there's much perf gain to be had given the (hopefully) random distribution of hashes |
| 13:05 | Modius | It would mean, for example, in a bulk add the root array (and rootward arrays) are only rebuilt once. The inserts split down the inserts into per-bucket sublists (it has not escaped me that *this* may make it a wash) |
| 17:31 | abrooks_ | jteo: re: Project Euler + Clojure -- truly the perfect match. I've been having fun with that pairing as well. |