From 4dde2625a1bcb5381b50b22e7212070266cd224f Mon Sep 17 00:00:00 2001 From: Sebastiano Tronto Date: Mon, 20 May 2024 12:52:57 +0200 Subject: Changed gendata h48 to bfs --- TODO.txt | 30 ++++++------------------------ 1 file changed, 6 insertions(+), 24 deletions(-) (limited to 'TODO.txt') diff --git a/TODO.txt b/TODO.txt index a93be99..07455b6 100644 --- a/TODO.txt +++ b/TODO.txt @@ -1,27 +1,9 @@ -In progress: go back to nissy-style BFS for eosep data computation - - (done) compute selfsim and cocsep representatives - - (done) implement set_eo_fast for invcoord_h48 - - (done) for invcoord_h48 remove erep, implement inverse esep coord - - (done) add unit tests for invcoord_h48 (compute tables, if needed) - - change to BFS - -TODO pruning tables: - - four different ruotines for k=4,2,1 for eoesep - - try: do not compute CO, but use its binary representation - (x8 memory for cocsep) - (isn't this already done?) - - numbers so far (eoesep h=0, k=0): - 0 1 - 1 1 - 2 4 - 3 34 - 4 331 - 5 3612 - 6 41605 - 7 474128 - 8 4953846 - 9 34776317 +TODO + - test h48 table by adding a parameter to stop computation at certain depth + - add back benchmark, test performance of table computation + - table generation is still slow, how to improve? profile! + selfsim could be a list of cubes (representatives) + - add h48 table for different values of k TODO checkdata (available from cube.h) and hash check for cocsep TODO benchmarks for solve and table generation -- cgit v1.3