aGrUM/pyAgrum 1.3.0 released

Posted on Mon 01 August 2022 in News

ANNOUNCE: aGrUM 1.3.0

aGrUM/pyAgrum 1.3.0 is out.

Changelog for 1.3.0

  • aGrUM

    • Fixing errors in doxygen configuration.
    • Adding a new class of discrete variable gum::NumericalDiscreteVariable whose support is a (finite) set of double.
    • Fixing new clang warnings.
    • Fixing a bug in gum::Potential::max() when all values are negative (thanks to Zakarie Aloui).
    • Adding a new method gum::learning::BNLearner::score() which computes the currently selected score for a node and its parents.
    • Adding a test for a variance being not negative in gum::ShaferSheoyLIMIDInference (thanks to Benjamin Datko).
    • Adding new methods gum::learning::BNLeaner::mutualInformation() and gum::learning::BNLearner::correctedMutualInformation().
  • pyAgrum

    • Adding a new class of discrete variable gum.NumericalDiscreteVariable whose support is a (finite) set of double.
    • Updating notebooks w.r.t. this new variable (16-Examples_quasiContinuous and 18-Examples-BayesianBetaCoin).
    • Adding a new method gum.BNLearner.score() which computes the currently selected score for a node and its parents.
    • Adding new methods gum.BNLeaner.mutualInformation() and gum.BNLearner.correctedMutualInformation().