Merge pull request #90 from IDSIA/dev

Release 0.2.1
IDSIA · Aug 28, 2021 · 1d98b0a · 1d98b0a
2 parents 29e4660 + 56327b1
commit 1d98b0a
Show file tree

Hide file tree

Showing 72 changed files with 2,320 additions and 600 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,3 +1,15 @@
+0.2.1
+===========
+
+- New *logics* Bayesian factors, such as `Not`, `And`, `Or`, and `NoisyOr` ( #86 )
+- New ApproxLP0 algorithm ( #83 )
+- Renamed `approxlp` package to `approxpl1`  ( #84 )
+- Added new `VertexToRandomBayesian` converter ( #84 )
+
+Extra:
+- Replaced CI/CD pipeline with GitHub actions ( #84 )
+- Added CITATION.cff and CITATION.bib files
+
 0.2.0
 ===========
 

diff --git a/README.md b/README.md
@@ -1,10 +1,9 @@
 [![GitHub version](https://badge.fury.io/gh/idsia%2Fcrema.svg)](https://badge.fury.io/gh/idsia%2Fcrema)
-[![Build Status](https://travis-ci.org/IDSIA/crema.svg?branch=master)](https://travis-ci.org/IDSIA/crema)
-
+![example workflow](https://github.com/IDSIA/crema/actions/workflows/maven.yaml/badge.svg?branch=master)
 
 <img src="docs/static/img/logo.png" alt="Crema" width="500"/>
 
-CreMA is a open-source java toolbox that provides multiple
+CreMA is an open-source java toolbox that provides multiple
 learning and inference algorithms for credal models.
 
 An example of exact inference in a credal network is given below.
@@ -72,7 +71,7 @@ Add the following code in the  pom.xml of your project:
         <dependency>
             <groupId>ch.idsia</groupId>
             <artifactId>crema</artifactId>
-            <version>0.2.0</version>
+            <version>0.2.1</version>
             <scope>compile</scope>
         </dependency>
     </dependencies>

diff --git a/docs/notes/bayesianinference.rst b/docs/notes/bayesianinference.rst
@@ -15,6 +15,7 @@ the model. This will return the posterior of a variable considered the root of t
 .. literalinclude:: ../../examples/BeliefPropagation.java
    :language: java
    :lines: 42-44
+   :dedent: 2
 
 
 To perform an inference on a variable, as an example if you want the marginal of ``P(A)``, use the ``query()`` method as
@@ -23,6 +24,7 @@ in the example below:
 .. literalinclude:: ../../examples/BeliefPropagation.java
    :language: java
    :lines: 55-60
+   :dedent: 2
 
 
 If you want to use evidence, you need to create first a ``TIntIntHashMap`` that will include the state of the various
@@ -31,6 +33,7 @@ variables, in the belo case  we query for ``P(A | B=0)``:
 .. literalinclude:: ../../examples/BeliefPropagation.java
    :language: java
    :lines: 62-69
+   :dedent: 2
 
 
 Full example:

diff --git a/docs/notes/contact.rst b/docs/notes/contact.rst
@@ -1,29 +1,20 @@
 Contact and Support
 ======================
 
-
-.. image:: ../_static/img/idsia.png
+.. image:: _static/img/idsia.png
    :scale: 60 %
    :align: center
-   :target: http://www.idsia.ch
-
+   :target: https://www.idsia.ch/
 
+Crema has been developed at the Swiss AI Lab IDSIA (Istituto Dalle Molle di Studi sull'Intelligenza Artificiale),
+a not-for-profit research institute for Artificial Intelligence.
 
-Crema has been developed at the Swiss AI Lab IDSIA (Istituto Dalle Molle di Studi sull'Intelligenza Artificiale).
 The members of the development and research team are:
 
-
 - David Huber ([email protected])
-
 - Rafael Cabañas ([email protected])
-
 - Alessandro Antonucci ([email protected])
-
 - Marco Zaffalon ([email protected])
-
 - Claudio Bonesana ([email protected])
 
-
 If you have any question, please use `Github issues <https://github.com/IDSIA/crema/issues>`_.
-
-
diff --git a/docs/notes/guideinference.rst b/docs/notes/guideinference.rst
@@ -7,17 +7,13 @@ run the query. The result is an object of class VertexFactor.
 
 .. literalinclude:: ../../examples/docs/inferEx1.java
    :language: java
-   :lines: 34-39
+   :lines: 39-42
 
 
 Approximate inference can be done by means of linear programming. For this, create
 the an object of class ``CredalApproxLP`` and then run the query. Note
 that the output is an ``IntervalFactor``.
 
-
-
 .. literalinclude:: ../../examples/docs/inferEx2.java
    :language: java
-   :lines: 48-54
-
-
+   :lines: 52-58
diff --git a/docs/notes/installation.rst b/docs/notes/installation.rst
@@ -16,7 +16,7 @@ Crema can be easily included at any maven project. For this, add the following c
         <dependency>
             <groupId>ch.idsia</groupId>
             <artifactId>crema</artifactId>
-            <version>0.1.6</version>
+            <version>0.2.1</version>
             <scope>compile</scope>
         </dependency>
     </dependencies>

diff --git a/docs/notes/markdownguide.md b/docs/notes/markdownguide.md
diff --git a/docs/static/img/idsia.png b/docs/static/img/idsia.png
diff --git a/docs/tutorials/domains.rst b/docs/tutorials/domains.rst
@@ -7,6 +7,7 @@ Domains
 .. contents:: Table of Contents
     :local:
 
+
 Domain interface
 =======
 
@@ -24,6 +25,7 @@ This simple interface declares basic methods to query the domain about variables
 .. Note::
     Returned arrays should never be modified!
 
+
 SimpleDomain
 ============
 
@@ -33,22 +35,24 @@ This class encapsulates two integer arrays. One with the variable labels and one
 .. literalinclude:: ../../src/test/java/ch/idsia/crema/tutorial/DomainsTutorial.java
     :start-after: [simple-domain]
     :end-before: [simple-domain]
-    :dedent: 8
+    :dedent: 2
 
 .. Warning::
     When creating a :code:`SimpleDomain` the list of variables must be sorted! 
     Crema will **not** automatically sort them, but for some operations will assume they are.
 
+
 DomainBuilder
 =============
-While creating a :code:`SimpleDomain` by passing the arrays of variables and their sizes is possible and valid, 
+
+While creating a :code:`SimpleDomain` by passing the arrays of variables and their sizes is possible and valid,
 a slightly more friendly method is available using the :code:`DomainBuilder`. 
 Laveraging the ellipses of Java the :code:`DomainBuilder` class avoids the explicit creation of the arrays as shown in the following example.
 
 .. literalinclude:: ../../src/test/java/ch/idsia/crema/tutorial/DomainsTutorial.java
     :start-after: [domain-builder-1]
     :end-before: [domain-builder-1]
-    :dedent: 8
+    :dedent: 2
 
 
 Strides
@@ -71,27 +75,26 @@ We we first look at how :code:`Strides` instances can be created conveniently.
 .. Note::
     The variable's cardinalities are accumlated starting from the variable at index 0.
 
-
 .. literalinclude:: ../../src/test/java/ch/idsia/crema/tutorial/DomainsTutorial.java
     :start-after: [strides]
     :end-before: [strides]
-    :dedent: 8
+    :dedent: 2
 
 Again, just as with the :code:`SimpleDomain`, creating the object specifying the arrays is valid, but not the most readable solution.
 The following example shows an alternative way of creation where variables are added along with their cardinality.
 
 .. literalinclude:: ../../src/test/java/ch/idsia/crema/tutorial/DomainsTutorial.java
     :start-after: [domain-builder-strides]
     :end-before: [domain-builder-strides]
-    :dedent: 8
+    :dedent: 2
 
 
 Alternative ways to create strides are based on operations on them. Generally Domains are considered unmutable objects and any alteration will result in a new instance.
 
 .. literalinclude:: ../../src/test/java/ch/idsia/crema/tutorial/DomainsTutorial.java
     :start-after: [strides-remove]
     :end-before: [strides-remove]
-    :dedent: 8
+    :dedent: 2
 
 A number of common set operations are available:
 
@@ -100,8 +103,6 @@ A number of common set operations are available:
 - remove
 
 .. _Enhanced domain:
+
 Working with Strides
 --------------------
-
-
-
diff --git a/docs/tutorials/factors.rst b/docs/tutorials/factors.rst
@@ -5,5 +5,5 @@
 Factors
 ===============
 
-Credo supports different ways to represent the probability functions defined over the variables. A structure of different categorization
-and abstraction around factors have been implemented. At the top of this all we have the concept of :code:`GenericFactor`. 
+Crema supports different ways to represent the probability functions defined over the variables. A structure of different categorization
+and abstraction around factors have been implemented. At the top of this all we have the concept of :code:`GenericFactor`.
diff --git a/docs/tutorials/networks.rst b/docs/tutorials/networks.rst
@@ -42,7 +42,7 @@ For a Bayesian Network we will use a :code:`BayesianFactor`.
 .. literalinclude:: ../../src/test/java/ch/idsia/crema/tutorial/NetworkTutorial.java
     :start-after: [creating-sparse-model]
     :end-before: [creating-sparse-model]
-    :dedent: 8
+    :dedent: 2
     :language: java
 
 

diff --git a/examples/CredalInferenceExample.java b/examples/CredalInferenceExample.java
@@ -11,8 +11,8 @@
 import ch.idsia.crema.factor.credal.linear.separate.SeparateHalfspaceFactorFactory;
 import ch.idsia.crema.factor.credal.vertex.separate.VertexFactor;
 import ch.idsia.crema.factor.credal.vertex.separate.VertexFactorFactory;
-import ch.idsia.crema.inference.approxlp.ApproxLP1;
-import ch.idsia.crema.inference.approxlp.CredalApproxLP;
+import ch.idsia.crema.inference.approxlp1.ApproxLP1;
+import ch.idsia.crema.inference.approxlp1.CredalApproxLP;
 import ch.idsia.crema.inference.approxlp2.ApproxLP2;
 import ch.idsia.crema.inference.ve.FactorVariableElimination;
 import ch.idsia.crema.inference.ve.VariableElimination;

diff --git a/examples/PGMpaper2.java b/examples/PGMpaper2.java
@@ -6,7 +6,7 @@
 import ch.idsia.crema.factor.credal.linear.interval.IntervalFactor;
 import ch.idsia.crema.factor.credal.linear.separate.SeparateHalfspaceFactor;
 import ch.idsia.crema.factor.credal.linear.separate.SeparateHalfspaceFactorFactory;
-import ch.idsia.crema.inference.approxlp.CredalApproxLP;
+import ch.idsia.crema.inference.approxlp1.CredalApproxLP;
 import ch.idsia.crema.model.graphical.DAGModel;
 import org.apache.commons.math3.optim.linear.Relationship;
 

diff --git a/examples/docs/inferEx2.java b/examples/docs/inferEx2.java
@@ -5,7 +5,7 @@
 import ch.idsia.crema.factor.credal.linear.interval.IntervalFactor;
 import ch.idsia.crema.factor.credal.linear.separate.SeparateHalfspaceFactor;
 import ch.idsia.crema.factor.credal.linear.separate.SeparateHalfspaceFactorFactory;
-import ch.idsia.crema.inference.approxlp.CredalApproxLP;
+import ch.idsia.crema.inference.approxlp1.CredalApproxLP;
 import ch.idsia.crema.model.graphical.DAGModel;
 import ch.idsia.crema.model.graphical.GraphicalModel;
 import org.apache.commons.math3.optim.linear.Relationship;