Skip to content

Releases: llnl/ibis

Release 1.2.1

Choose a tag to compare

@olson59 olson59 released this 27 Feb 17:29

Fixed

  • The PCE model is more accurate now that we're using the global max, and other modifications to make it more robust.

Release 1.2.0

Choose a tag to compare

@olson59 olson59 released this 17 Sep 05:27

Description

This release features new plots and some bug fixes.

New in this release

Sensitivity methods new score and rank plots

  • One-at-a-time: calculates each parameters individual effect shows the magnitude in a score plot, and compares features in a rank plot. ibis.sensitivity.oat_score_plot and ibis.sensitivity.oat_rank_plot
  • Sobol Indices: Calculates contributions of individual parameters and their interactions. The score plot shows first order or total order sensitivity indices. The rank plot compares them across the outputs. ibis.sensitivity.sobol_score_plot and ibis.sensitivity.sobol_rank_plot
  • Morris method: Calculates elementary effects for each path through the parameter space, and calculates statistics mu* and sigma across all the trajectories to understand main effects and interactions. Score plot can show mu* and sigma, and rank plot just compares parameters based on mu* or sigma. ibis.sensitivity.morris_score_plot and ibis.sensitivity.morris_rank_plot
  • These plots are also available as Kosh operators from ibis.kosh_operators.KoshSensitivityPlots

Bug fixes

  • Updated some indexing and variable assignment methods to be compatible with numpy > 2.0
  • Other fixes and updates to the testing suite

Release 1.1.1

Choose a tag to compare

@olson59 olson59 released this 31 Mar 22:49

1.1.1 Release

Description

This release includes some bug fixes and an update to the basic descriptions.

Improvements

  • Updated descriptions for the main modules
  • More explanation for discrepancy MCMC sampling

Bug fixes

  • KoshMCMC function only has experimental data coming from a Kosh store.
  • Fixed loop to allow for any number of quantity of interest and experiments in the KoshMCMC function.
  • Allow for the case of only one quantity of interest (QOI) in the sensitivity plot, variance_network_plot, and rank plot.

Release 1.1.0

Choose a tag to compare

@olson59 olson59 released this 12 Aug 22:53

1.1.0 Release

This release introduces some new features

New in this release

Added Kosh operators to Ibis to be able to use
IBIS UQ and sensitivity methods with Kosh datasets.

  • KoshMCMC
  • KoshOneAtATimeEffects
  • KoshSensitivityPlots

A sobol_indices function has been added to the sensitivity module. It's meant
to be used with the SobolIndexSampler in the Trata sampler module or similar.