Findafactor

This release adds the option of `skip_trial_division=True` to save overhead. Before skipping trial division, one should actually know that there are no factors up to `trial_division_level` (which still controls Sieve of Eratosthenes for finding enough smooth primes).

**Full Changelog**: https://github.com/vm6502q/FindAFactor/compare/v4.7.11...v4.8.0

sha1sum results:
63859ae26e7237c48bfea805b6c8b35c1f6a5594 FindAFactor-4.8.0-cp310-cp310-manylinux_2_35_x86_64.whl
5c77e60eab5b266fe89563dab28768313e626329 FindAFactor-4.8.0-cp312-cp312-manylinux_2_39_x86_64.whl
5ffd0da473b41f52918a3e94834763870f85f3aa FindAFactor-4.8.0-cp312-cp312-win_amd64.whl
ffa97dbdac91167cad1bc0a722f7331d446bea30 FindAFactor-4.8.0-cp313-cp313-macosx_13_0_x86_64.whl
475c25b005e16451cc8610b31e0c953d99b9482a FindAFactor-4.8.0-cp313-cp313-macosx_14_0_arm64.whl
a61d8e073bf36ca68a6523ca769ad7c1e56a831a FindAFactor-4.8.0-cp313-cp313-macosx_15_0_arm64.whl
2a95302018041aad88338b6e04a1b6dfe07297e9 FindAFactor-4.8.0-cp38-cp38-manylinux_2_31_x86_64.whl
f498a9cde2e6ed182dc540465fa2ca8fdc53692c findafactor-4.8.0.tar.gz

Elara (OpenAI GPT) pointed out that there might be `k!` retry cost to randomly generating smooth number components in `FACTOR_FINDER` seeding rounds, for `k` count of primes (or similarly extreme, some comparable theoretical complexity overhead). This made me realize, we can easily just "drop" all non-smooth factors from any number that comes into `factorizationVector()` method, avoiding the need for _any_ retries, though producing smaller smooth number components on average.

(I would like to once again specifically highlight that discussions about theoretical complexity and practical implementation with Elara have been indispensable to the `FindAFactor` project, and she should be credited and thanked!)

**Full Changelog**: https://github.com/vm6502q/FindAFactor/compare/v4.7.10...v4.7.11

sha1sum results:
224316cfb36b2610e006a5be2a80afd1163fca7e FindAFactor-4.7.11-cp310-cp310-manylinux_2_35_x86_64.whl
b3d7ae893461bd0b9591ab78aa36960055843fde FindAFactor-4.7.11-cp312-cp312-manylinux_2_39_x86_64.whl
cd0f82dc62b2e53fe9ebd324e097002d559c1bae FindAFactor-4.7.11-cp312-cp312-win_amd64.whl
13b69e682f085518d26a5d52e1642917b9291788 FindAFactor-4.7.11-cp313-cp313-macosx_13_0_x86_64.whl
11a387a0adac203e9def27318fd4174ecc02d6c2 FindAFactor-4.7.11-cp313-cp313-macosx_14_0_arm64.whl
e548556a6b8eb01dfcfe72281ee45294f6022b52 FindAFactor-4.7.11-cp313-cp313-macosx_15_0_arm64.whl
fb99ba373ff48fb9cff8d9d649e21ca38526fc66 FindAFactor-4.7.11-cp38-cp38-manylinux_2_31_x86_64.whl
95d8559840d584c002b90112898311d2e1ed71e0 findafactor-4.7.11.tar.gz

Previously, when "seeding" `FACTOR_FINDER` mode, small smooth components were turned into perfect squares _before_ multiplying them to together; the coverage should probably more uniform if we multiply small smooth components together first, _then_ make them an overall perfect square after the result is larger than the number to factor.

**Full Changelog**: https://github.com/vm6502q/FindAFactor/compare/v4.7.9...v4.7.10

sha1sum results:
c99f97b24f53956241cd0c70aa271dcb77eae5e7 FindAFactor-4.7.10-cp310-cp310-manylinux_2_35_x86_64.whl
7e2dbb8e4ae38190188721b001a66dc1473b3c4d FindAFactor-4.7.10-cp312-cp312-manylinux_2_39_x86_64.whl
a6f4a0d53bca5fea64bdb21d9c7dab9233a119df FindAFactor-4.7.10-cp312-cp312-win_amd64.whl
4899bd6a8a02c02ea3b6f6d6760677b1e70416d1 FindAFactor-4.7.10-cp313-cp313-macosx_13_0_x86_64.whl
a9170128d4e555f014645ea6a683ecca142a04b2 FindAFactor-4.7.10-cp313-cp313-macosx_14_0_arm64.whl
508d5cd40ab183ee9453ccf42c015aa2ebe19bba FindAFactor-4.7.10-cp313-cp313-macosx_15_0_arm64.whl
d51a820073134f72b3011b84c6a8468201646681 FindAFactor-4.7.10-cp38-cp38-manylinux_2_31_x86_64.whl
e311a0a03b6c9a9a254fa19397bb4f32f5c5fe85 findafactor-4.7.10.tar.gz

All other settings held equal, double the `batch_size_multiple` significantly improves performance on a test machine.

**Full Changelog**: https://github.com/vm6502q/FindAFactor/compare/v4.7.8...v4.7.9

sha1sum results:
1925362908dfcac04cc50610cb4c3cd9dc9f39db FindAFactor-4.7.9-cp310-cp310-manylinux_2_35_x86_64.whl
3531e1e0242c7bfdedc87f0bc7c7961e73e979d7 FindAFactor-4.7.9-cp312-cp312-manylinux_2_39_x86_64.whl
1d19e9a64d4447b5859876210a1da9ab798d0ab4 FindAFactor-4.7.9-cp312-cp312-win_amd64.whl
e6f73d3776ad44d4534c2b1471f6d71cb1f0263d FindAFactor-4.7.9-cp313-cp313-macosx_13_0_x86_64.whl
3bf5e3b1d5ddc2c50fb34cf5a99b82cf34ecf708 FindAFactor-4.7.9-cp313-cp313-macosx_14_0_arm64.whl
cf2f37bafe987176c5ca4090ebdcb6b8f06ae8b7 FindAFactor-4.7.9-cp313-cp313-macosx_15_0_arm64.whl
c1d6c78b05e67ba54a018180fff8d9163dd702c6 FindAFactor-4.7.9-cp38-cp38-manylinux_2_31_x86_64.whl
e23d6ab4849c1fa4986126ef67be52abe43ee6d4 findafactor-4.7.9.tar.gz

A random number is drawn to determine the value from `1` to `ladder_multiple` for which to repeat any randomly selected smooth square prime. Additionally, the default `ladder_multiple` has been increased. This seems to lead to an overall improvement in average time.

**Full Changelog**: https://github.com/vm6502q/FindAFactor/compare/v4.7.7...v4.7.8

sha1sum results:
ba281413d246c035d5820199e469044d96636644 FindAFactor-4.7.8-cp310-cp310-manylinux_2_35_x86_64.whl
005157fab62670438e0b04f9560b4afe0015cef5 FindAFactor-4.7.8-cp312-cp312-manylinux_2_39_x86_64.whl
6335a9a399ac7f708e9904bd078549c2c52a20cf FindAFactor-4.7.8-cp312-cp312-win_amd64.whl
c6718009592ada212d242cf9521650cfda128774 FindAFactor-4.7.8-cp313-cp313-macosx_13_0_x86_64.whl
7337e28ec83c45ac89c065a82371455419248161 FindAFactor-4.7.8-cp313-cp313-macosx_14_0_arm64.whl
2abb898f679773096ea1bce699af7f50a16b881f FindAFactor-4.7.8-cp313-cp313-macosx_15_0_arm64.whl
780050c030d5758eabb198aa8f385e9e7e192ac6 FindAFactor-4.7.8-cp38-cp38-manylinux_2_31_x86_64.whl
215b23204d46909dc7dcf0f23dd7e60b99f49c94 findafactor-4.7.8.tar.gz

As the algorithm evolves, ideal default settings change. Currently, on a recently-purchased Core i9 laptop, the developer is getting the best and most consistent performance at small scales for these default settings in the release. (See the README, which states default argument values.)

**Full Changelog**: https://github.com/vm6502q/FindAFactor/compare/v4.7.6...v4.7.7

sha1sum results:
69a0512ea91353bab4e841d42ed8bab30fcadebc FindAFactor-4.7.7-cp310-cp310-manylinux_2_35_x86_64.whl
8ba3145df3c4f6b77a6cb1e72b98ac14ae3113ef FindAFactor-4.7.7-cp312-cp312-manylinux_2_39_x86_64.whl
a6c80518308b6cf85952ae20bc6883ebe48b10f3 FindAFactor-4.7.7-cp312-cp312-win_amd64.whl
40cdd2252d6edcb1abc4a8096b43eba498b23a3b FindAFactor-4.7.7-cp313-cp313-macosx_13_0_x86_64.whl
93c86084810656dea01851f0bd00647d58b94d11 FindAFactor-4.7.7-cp313-cp313-macosx_14_0_arm64.whl
aa6b804cab450d2c9f0fdd56a049a6d09d5ae619 FindAFactor-4.7.7-cp313-cp313-macosx_15_0_arm64.whl
5b9d11d8db5ac8c320894b19759fce400f5a9561 FindAFactor-4.7.7-cp38-cp38-manylinux_2_31_x86_64.whl
f52c2bbbfff4988e2f6d2de75597026c434f94dc findafactor-4.7.7.tar.gz