Hi,
Nice work! I was just checking, and https://github.com/meelgroup/cmsgen seems to be missing from the samplers. Under normal circumstances it should work quite a bit better than other other uniform-like (but not guaranteed uniform) samplers. See our paper here Basically, it's a sampler that was born out of Barbarik. It can fool Barbarik relatively easily, unlike other uniform-like samplers. In fact, due to Barbarik's failure to detect, we came up with ScalBarbarik: https://github.com/meelgroup/scalbarbarik/ (paper here) to reliably detect it :)
Anyway, just wanted to say that in case you have quicksampler, maybe you want to have CMSGen -- it's supposed to be quite a bit quicker and should produce better quality samples. Check it out :)
Cheers, and let me know what you think,
Mate (one of the authors of CMSGen and ScalBarbarik)
(pinging @kuldeepmeel in case he's interested)
Hi,
Nice work! I was just checking, and https://github.com/meelgroup/cmsgen seems to be missing from the samplers. Under normal circumstances it should work quite a bit better than other other uniform-like (but not guaranteed uniform) samplers. See our paper here Basically, it's a sampler that was born out of Barbarik. It can fool Barbarik relatively easily, unlike other uniform-like samplers. In fact, due to Barbarik's failure to detect, we came up with ScalBarbarik: https://github.com/meelgroup/scalbarbarik/ (paper here) to reliably detect it :)
Anyway, just wanted to say that in case you have
quicksampler, maybe you want to have CMSGen -- it's supposed to be quite a bit quicker and should produce better quality samples. Check it out :)Cheers, and let me know what you think,
Mate (one of the authors of CMSGen and ScalBarbarik)
(pinging @kuldeepmeel in case he's interested)