Issues - trueskill-tt - Gitea: Git with a cup of tea

logaritmisk/trueskill-tt

Labels Milestones New Issue

The joint precision matrix is 0.19% dense and we factorise it densely enhancement numerics performance

by logaritmisk was closed 2026-09-10 05:08:22 +00:00

Cache the joint factorisation: repeat posterior_of queries re-pay the full O(n^3) solve enhancement performance

by logaritmisk was closed 2026-09-08 08:24:46 +00:00

erfc is a 1.2e-7 approximation and sets the accuracy floor for the whole crate numerics performance

by logaritmisk was closed 2026-09-07 20:25:11 +00:00

Ingestion discards all convergence state, so appending one event costs a full re-converge performance

by logaritmisk was closed 2026-08-28 06:01:55 +00:00

Per-slice SkillStore is dense over the global index space — memory scales with the largest competitor index, not slice size performance

by logaritmisk was closed 2026-08-27 16:02:16 +00:00

learning_curves() is O(n²) because KeyTable::key is a linear scan performance

by logaritmisk was closed 2026-08-04 20:13:43 +00:00

Hot-path Gaussian arithmetic round-trips through sqrt; convergence deltas computed in moment space numerics performance

by logaritmisk was closed 2026-08-04 20:13:42 +00:00

Adding events one at a time is quadratic — 130× slower than batching at n=2000 performance

by logaritmisk was closed 2026-08-04 20:13:40 +00:00

Decide whether rayon becomes a default-on feature blocked decision performance

by logaritmisk was closed 2026-09-08 13:53:48 +00:00

Cross-slice parallelism: per-slice dirty bits to skip clean slices enhancement performance

by logaritmisk was closed 2026-08-27 21:50:46 +00:00