0.1.1

Author: idc9

Readme.md

@@ -36,7 +36,7 @@ The built in cross-validation functionality supports
 - custom evaluation metrics
 - custom selection rules such as the '1se' rule from the glmnet package
 
-We provide a built in FISTA algorithm [(Beck and Teboulle, 2009)](https://epubs.siam.org/doi/pdf/10.1137/080716542?casa_token=cjyK5OxcbSoAAAAA:lQOp0YAVKIOv2-vgGUd_YrnZC9VhbgWvZgj4UPbgfw8I7NV44K82vbIu0oz2-xAACBz9k0Lclw) that covers most glm loss + non-smooth penalty combinations (`ya_glm.opt` is inspired by [pyunlocbox](https://github.com/epfl-lts2/pyunlocbox) and [lightning](https://github.com/scikit-learn-contrib/lightning)). **It is straightforward for you to plug in your favorite penalized GLM optimization algorithm.**
+We provide a built in FISTA algorithm ([Beck and Teboulle, 2009](https://epubs.siam.org/doi/pdf/10.1137/080716542?casa_token=cjyK5OxcbSoAAAAA:lQOp0YAVKIOv2-vgGUd_YrnZC9VhbgWvZgj4UPbgfw8I7NV44K82vbIu0oz2-xAACBz9k0Lclw)) that covers most glm loss + non-smooth penalty combinations (`ya_glm.opt` is inspired by [pyunlocbox](https://github.com/epfl-lts2/pyunlocbox) and [lightning](https://github.com/scikit-learn-contrib/lightning)). **It is straightforward for you to plug in your favorite penalized GLM optimization algorithm.**
 
 We aim to add additional loss functions (e.g. gamma, cox regression) and penalties (e.g. generalized Lasso, TV1)
 

setup.py

@@ -15,7 +15,7 @@
 
 
 setup(name='ya_glm',
-      version='0.1.0',
+      version='0.1.1',
       description='A flexible package for fitting penalized generalized linear models.',
       author='Iain Carmichael',
       author_email='idc9@cornell.edu',