DOC add link to arxiv paper, update readme (#1)

Co-authored-by: mathurinm <[email protected]>

Author: QB3

.circleci/config.yml

@@ -24,7 +24,7 @@ jobs:
                 echo "Merging $(cat merge.txt)";
                 git remote add upstream https://github.com/mathurinm/skglm.git;
                 git pull --ff-only upstream "refs/pull/$(cat merge.txt)/merge";
-                git fetch upstream master;
+                git fetch upstream main;
               fi
 
         # If both keys are in the same command only one is restored

README.rst

@@ -1,7 +1,10 @@
 skglm
-========
+=====
 
-``skglm`` is a library that provide better sparse generalized linear model for scikit-learn.
+|image0|
+
+
+``skglm`` is a library that provide better sparse generalized linear models for scikit-learn.
 Its main features are:
 
 - **speed**: problems with millions of features can be solved in seconds. Default solvers rely on efficient coordinate descent with numba just in time compilation.
@@ -10,24 +13,35 @@ Its main features are:
 - **scope**: support for many missing models in scikit-learn - weighted Lasso, arbitrary group penalties, non convex sparse penalties, etc.
 
 
-Cite
-----
+Currently, the package handles any combination of the following datafits:
 
-If you use this code, please cite
+- quadratic
+- logistic loss
+- multitask quadratic
 
-.. code-block:: none
+and the following penalties:
 
-    @online{skglm,
-        title={Beyond L1: Faster and Better Sparse Models with skglm},
-        author={Q. Bertrand and Q. Klopfenstein and P.-A. Bannier and G. Gidel and M. Massias},
-        year={2022},
-        url={https://arxiv.org/abs/2204.07826}
-    }
+- L1 norm
+- weighted L1 norm
+- L1 + L2 squared norm (elastic net)
+- MCP
+- L05 and L2/3 penalties
+
+
+The estimators follow the scikit-learn API, come with automated parallel cross-validation, and support both sparse and dense data.
+
+.. with optionally feature centering, normalization, and unpenalized intercept fitting.
+
+Documentation
+=============
+
+Please visit https://mathurinm.github.io/skglm/ for the latest version
+of the documentation.
 
 
+Install and work with the development version
+=============================================
 
-Installing the development version
-----------------------------------
 First clone the repository available at https://github.com/mathurinm/skglm::
 
     $ git clone https://github.com/mathurinm/skglm.git
@@ -41,4 +55,44 @@ To check if everything worked fine, you can do::
 
     $ python -c 'import skglm'
 
-and it should not give any error message.
+and it should not give any error message.
+
+
+
+Demos & Examples
+================
+
+In the `example section <https://mathurinm.github.io/skglm/auto_examples/index.html>`__ of the documentation,
+you will find numerous examples on real life datasets,
+timing comparison with other estimators, easy and fast ways to perform cross validation, etc.
+
+
+Dependencies
+============
+
+All dependencies are in the ``./requirements.txt`` file.
+They are installed automatically when ``pip install -e .`` is run.
+
+Cite
+----
+
+If you use this code, please cite
+
+.. code-block:: none
+
+    @online{skglm,
+        title={Beyond L1: Faster and Better Sparse Models with skglm},
+        author={Q. Bertrand and Q. Klopfenstein and P.-A. Bannier and G. Gidel and M. Massias},
+        year={2022},
+        url={https://arxiv.org/abs/2204.07826}
+    }
+
+
+
+
+ArXiv links:
+
+- https://arxiv.org/pdf/2204.07826.pdf
+
+.. |image0| image:: https://github.com/mathurinm/skglm/workflows/pytest/badge.svg
+   :target: https://github.com/mathurinm/skglm/actions?query=workflow%3Abuild

doc/index.rst

@@ -25,16 +25,17 @@ If you use this code, please cite
     @online{skglm,
         title={Beyond L1 norm with skglm},
         author={Q. Bertrand and Q. Klopfenstein and P.-A. Bannier and G. Gidel and M. Massias},
-        booktitle={in preparation},
+        journal = {arXiv preprint arXiv:2204.07826},
+        url={https://arxiv.org/pdf/2204.07826.pdf}
         year={2022},
     }
 
 
-Installing the latest release
------------------------------
-Run::
+.. Installing the latest release
+.. -----------------------------
+.. Run::
 
-    $ pip install scikit-contrib-skglm
+..    $ pip install scikit-contrib-skglm
 
 
 Installing the development version