Merge pull request #5 from edahelsinki/docs

Badges

Author: Aggrathon

.github/workflows/python-docs.yml

@@ -1,12 +1,14 @@
 # This workflow will build the documentation using mkdocs and push it to the gh-pages branch when a release is created
 # For more information see: https://www.mkdocs.org/user-guide/deploying-your-docs/
 
-name: Build documentation
+name: docs
+
 on:
   workflow_dispatch:
   release:
     branches: [master, main]
     types: [released]
+
 jobs:
   deploy:
     runs-on: ubuntu-latest
@@ -15,9 +17,11 @@ jobs:
     - uses: actions/setup-python@v2
       with:
         python-version: 3.x
+
     - name: Install dependencies
       run:  |
         python -m pip install --upgrade pip
         python -m pip install -r docs/requirements-dev.txt
+        
     - name: Build and deploy
       run: mkdocs gh-deploy --force --clean --verbose

.github/workflows/python-publish.yml

@@ -6,7 +6,7 @@
 # separate terms of service, privacy policy, and support
 # documentation.
 
-name: Upload Python Package
+name: publish
 
 on:
   release:

.github/workflows/python-pytest.yml

@@ -1,7 +1,7 @@
 # This workflow will install Python dependencies, run tests and lint with a variety of Python versions
 # For more information see: https://help.github.com/actions/language-and-framework-guides/using-python-with-github-actions
 
-name: Test Python Package
+name: tests
 
 on:
   push:

README.md

@@ -1,9 +1,16 @@
-![PySLISE Banner Image](docs/pyslise_banner.webp)
+![PySLISE Banner Image](docs/pyslise_banner.webp)  
+[![PyPI](https://img.shields.io/pypi/v/slise)](https://pypi.org/project/slise/)
+[![Documentation](https://github.com/edahelsinki/pyslise/actions/workflows/python-docs.yml/badge.svg)](https://edahelsinki.github.io/pyslise/docs/slise/)
+[![Tests](https://github.com/edahelsinki/pyslise/actions/workflows/python-pytest.yml/badge.svg)](https://github.com/edahelsinki/pyslise/actions/workflows/python-pytest.yml)
+[![GitHub license](https://img.shields.io/github/license/edahelsinki/pyslise)](https://github.com/edahelsinki/pyslise/blob/master/LICENSE)
+[![Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/gh/edahelsinki/pyslise/HEAD?labpath=examples)
+
 # SLISE - Sparse Linear Subset Explanations
 
 Python implementation of the SLISE algorithm. The SLISE algorithm can be used for both robust regression and to explain outcomes from black box models.
-For more details see [the original paper](https://rdcu.be/bVbda) or the [robust regression paper](https://rdcu.be/cFRHD).
+For more details see [the original paper](https://rdcu.be/bVbda) or [the robust regression paper](https://rdcu.be/cFRHD).
 Alternatively for a more informal overview see [the presentation](https://github.com/edahelsinki/slise/raw/master/vignettes/presentation.pdf), or [the poster](https://github.com/edahelsinki/slise/raw/master/vignettes/poster.pdf).
+Finally, for learning to use the python package there are several [examples](https://github.com/edahelsinki/pyslise/tree/master/examples/) and [the documentation](https://edahelsinki.github.io/pyslise/docs/slise/).
 
 > *Björklund A., Henelius A., Oikarinen E., Kallonen K., Puolamäki K.* (2019)  
 > **Sparse Robust Regression for Explaining Classifiers.**  

examples/README.md

@@ -0,0 +1,7 @@
+# Examples
+
+This directory contains jupyter notebooks that demonstrate how to use SLISE.
+The first notebook focuses on how to use SLISE for robust regression and why robust regression can be advantageous over normal linear regression.
+The following notebooks demonstrates how SLISE can be used to explain outcomes from black box models on various datasets.
+
+The notebooks can be viewed directly on GitHub by opening them. But they can also be viewed interactively through binder: [![Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/gh/edahelsinki/pyslise/HEAD?labpath=examples) (note that the binder instances tend to be quite slow).

examples/requirements.txt

@@ -0,0 +1,6 @@
+tensorflow
+wordcloud
+nltk
+slise
+tbb
+icc_rt

mkdocs.yml

@@ -8,6 +8,7 @@ theme:
   icon:
     logo: material/book-open
   favicon: pyslise_icon.webp
+  font: false
   palette:
     - media: "(prefers-color-scheme: light)"
       scheme: default

setup.cfg

@@ -1,6 +1,6 @@
 [metadata]
 name = slise
-version = 2.1.2
+version = 2.1.3
 author = Anton Björklund
 author_email = anton.bjorklund@helsinki.fi
 description = The SLISE algorithm for robust regression and explanations of black box models

slise/initialisation.py

@@ -199,7 +199,7 @@ def initialise_candidates(
     beta_max: float = 20,
     max_approx: float = 1.15,
     pca_treshold: int = 10,
-    num_init: int = 500,
+    num_init: Optional[int] = None,
     max_iterations: int = 300,
     beta_max_init: float = 2.5,
     min_beta_step: float = 1e-8,
@@ -224,6 +224,8 @@ def initialise_candidates(
     Returns:
         Tuple[np.ndarray, float]: `(alpha, beta)`.
     """
+    if num_init is None:
+        num_init = min(500, 3 * 4 ** X.shape[1])
     # Prepare parameters
     epsilon = epsilon**2
     beta_max = min(beta_max, beta_max_init) / epsilon
@@ -273,7 +275,7 @@ def initialise_candidates2(
     weight: Optional[np.ndarray] = None,
     beta_max: float = 20,
     max_approx: float = 1.15,
-    num_init: int = 500,
+    num_init: Optional[int] = None,
     max_iterations: int = 300,
     beta_max_init: float = 2.5,
     min_beta_step: float = 1e-8,
@@ -297,6 +299,8 @@ def initialise_candidates2(
     Returns:
         Tuple[np.ndarray, float]: `(alpha, beta)`.
     """
+    if num_init is None:
+        num_init = min(500, 3 * 4 ** X.shape[1])
     # Prepare parameters
     epsilon = epsilon**2
     beta_max = min(beta_max, beta_max_init) / epsilon