Master and v1.0.4 runecos.exe return different results
Closed this issue · 1 comments
chenhappy commented
when I printed first four elements of best_x in runecos.c demo, master and v1.0.4 have different solutions. Master: -0.2919 3.8371 5.4935 0.7858 v1.0.4: -0.6160 10.1737 18.8111 1.0115. which one is correct? when I check some sum(A[1,]*best_x) neither of them equal to b[1] in data.h. I must have missed something. In any case should we make Master 'to-be-released' version, at least all the tests are correct? Thanks.
chenhappy commented
sorry just realized that the solution should be in x not best_x.