ctak/include/weights.c, branch action_list_caching An implementation of Tak and a computer opponent in C https://git.l-3.space/ctak/atom?h=action_list_caching 2026-08-28T18:37:41Z Rename ct_k -> ct, IANAL but ... 2026-08-28T18:37:41Z tslil tslil@posteo.de 2021-02-05T21:49:14Z urn:sha1:ce09b6d94ac3dbf426e959b14fd7731001a003d1 Just some #weightgoals ;) 2026-08-28T18:37:41Z tslil clingman tslil@posteo.de 2021-02-01T04:33:39Z urn:sha1:328c8d1e3094a942d6a2edd933c9cc4ab09daab1 It turns out that while i was training on a 0/1 classification problem, i was using 2*eval - 1. Training using this function instead, and on bot-dominated game choices (chosen_player in extract.sh) seems to have given a better evaluation function. At the least, Morten's swindle doesn't work anymore. Added license information! 2026-08-28T18:37:41Z tslil clingman tslil@posteo.de 2021-01-29T02:30:30Z urn:sha1:a77062ce3dc9a99ad0662010cc7eeb00dd5036a6 Finally fixed memory order of weights! 2026-08-28T18:37:41Z tslil clingman tslil@posteo.de 2021-01-18T16:44:50Z urn:sha1:c1750c0cb93a7b4ba70e1342eaa7c62e659f1dc5 New weights from a wider training 2026-08-28T18:37:41Z tslil clingman tslil@posteo.de 2021-01-17T06:19:00Z urn:sha1:571afc6c72709910636979e20c0dd82dd0c66a46 Went back up to 1986 weights 2026-08-28T18:37:41Z tslil tslil@posteo.de 2021-01-16T19:08:46Z urn:sha1:50b9955b040005e489127a11cf9c1648438a6266 There was a bug in the minimax! 2026-08-28T18:37:41Z tslil tslil@posteo.de 2021-01-15T17:34:44Z urn:sha1:4ec810a2bb0ed393e59e0a03f8c98ec09805e989 Better trained weights, an important oversight in minimax (steps <=) 2026-08-28T18:37:41Z tslil tslil@posteo.de 2021-01-14T20:09:39Z urn:sha1:c1773e2712c5878e5f40e47e4a463a14724f55a8 I still can't tell whether if ((ply & 1) == 0) this = this - 1.0; is correct. Seems tolerable, but not great. Still wont win or lose ??? 2026-08-28T18:37:41Z tslil tslil@posteo.de 2021-01-14T06:05:45Z urn:sha1:b75526290b97e26d3fc3bbefef79f50a7febfe7d Fixed minimax (!), fixed bugs in tak.c 2026-08-28T18:37:41Z tslil tslil@posteo.de 2021-01-14T04:41:32Z urn:sha1:cf7d665598e34e32d258775b583968e758c8f2fb With minimax of depth 1 the evaluation function seems alright with the current method of training and generating weights