2
0
mirror of https://github.com/carlostrub/sisyphus synced 2024-10-31 09:20:15 +00:00
Commit Graph

16 Commits

Author SHA1 Message Date
Carlo Strub
9376143894 please gometalinter 2018-04-22 21:36:57 +02:00
Carlo Strub
54af109661
if word list is too long, only take a random subsample (fixes #11) 2018-04-05 22:30:18 -03:00
Carlo Strub
e7ac26f264
Add dry run mode that does not move mails after classification (fixes #8) 2018-02-03 22:08:12 +01:00
Carlo Strub
3a95f5cf1b
Use path/filepath for cleaner and safer path generation 2018-01-11 21:35:27 +01:00
Carlo Strub
cb1d924333 please gometalinter 2017-09-17 19:17:43 +02:00
Carlo Strub
bbee193679 improve memory footprint 2017-09-17 00:56:17 +02:00
Carlo Strub
87b51ebee0 glide up. load mails correctly before learning or classifying them.
improve some logging messages
2017-06-05 13:33:32 +00:00
Carlo Strub
3c665ed446 classify also if nothing have ever been learned 2017-05-28 20:53:43 +00:00
Carlo Strub
5413fdb4d4 Do not calculate harmonic mean if there is no probability -- otherwise
it crashes
2017-05-27 20:32:23 +00:00
Carlo Strub
1e67467307 use logrus properly (fixes #1) 2017-05-23 20:43:17 +00:00
Carlo Strub
04cc804722 replace log by logrus 2017-05-18 19:21:46 +00:00
Carlo Strub
276e665172 wrong prior calculation 2017-05-15 22:26:44 +00:00
Carlo Strub
36a3cb6374 make gometalinter happier 2017-05-13 22:34:54 +00:00
Carlo Strub
f76bfbd6eb add classify test and fix other tests 2017-05-13 21:22:23 +00:00
Carlo Strub
639ffadfe3 fix some static code errors 2017-05-10 17:23:25 +00:00
Carlo Strub
b5b3792efe implement learning and classifying using Bayes' rule and Hyperloglog
data structures -- still way to go though.
2017-05-10 03:45:11 +00:00