libgenetic.git/src, branch main Genetic Althorithm (GA) foundation library. https://git.xagasoft.com/Smaller Projects/libgenetic.git/atom?h=main 2019-08-08T22:30:47+00:00 More features. 2019-08-08T22:30:47+00:00 Mike Buland eichlan@xagasoft.com 2019-08-08T22:30:47+00:00 urn:sha1:042038dff89fa3a63e6613ade4540c7347cab953 Added more features to the simulation. 2019-08-08T16:59:41+00:00 Mike Buland eichlan@xagasoft.com 2019-08-08T16:59:41+00:00 urn:sha1:ca9f059a4295db6269728b0b747036f995c777a9 Added I/O test for binary phenotypes. 2012-08-05T07:27:43+00:00 Mike Buland mike@xagasoft.com 2012-08-05T07:27:43+00:00 urn:sha1:e9cfcc6dfc8427904ab227eea2452ccddd8f295c Added option to recalculate survivors. 2012-08-05T07:26:45+00:00 Mike Buland mike@xagasoft.com 2012-08-05T07:26:45+00:00 urn:sha1:8a0d33b5ebb3497d158f2d5c050c69be3fcd1e0d This accomidates simulations where the test conditions can change as the simulation runs, which would effect the scores of phenotypes that have already been tested. Made some changes to make serializing workable. 2012-08-05T00:34:57+00:00 Mike Buland mike@xagasoft.com 2012-08-05T00:34:57+00:00 urn:sha1:2406848173c445a94a1710106116ad796a8bacb9 There's no official format yet, but that's coming, most likely. Optomized the mating / mutation setup. 2012-08-04T22:59:49+00:00 Mike Buland mike@xagasoft.com 2012-08-04T22:59:49+00:00 urn:sha1:aaa733d789612289396d69bb6cf3a0e5de1798a8 It should work the same, but it selects the number of bits to mutate first, then selects the individual bits instead of testing a random number on each bit. For very large phenotypes this will be significantly faster while producing about the same result. New commit method 2012-08-02T21:20:58+00:00 Mike Buland mike@xagasoft.com 2012-08-02T21:20:58+00:00 urn:sha1:3cdf68e08d438a0ced0f5f8d957a6aafd68cb43a Persistant multi-threading works perfectly. 2012-07-18T07:26:34+00:00 Mike Buland mike@xagasoft.com 2012-07-18T07:26:34+00:00 urn:sha1:299640ce5f6499e07c5799f48897ac5a77e72c54 It uses the forever-blocking variation of the Bu::SynchroQueue dequeue to maximize efficiency. It all works multi-threaded. 2012-07-17T15:11:40+00:00 Mike Buland mike@xagasoft.com 2012-07-17T15:11:40+00:00 urn:sha1:a2679e3bc2f407fae6b97908aeebb215c6678ebc It restarts the threads every generation. Maybe not the worst thing every, but it seems like it would be better to stay in the thread and notify the main thread that they're done with that iteration. Simple saving/loading - consider this temporary. 2012-07-09T21:34:11+00:00 Mike Buland mike@xagasoft.com 2012-07-09T21:34:11+00:00 urn:sha1:a48556c1c5c01892cf493cd3eb4ff969a2f548fd It should be a little more compact, and have more header info so we can reconstruct the entire object from the file, right now you have to know the size in bits.