What's Changed

• Heavily revised and improved documentation.
• A range of fixes and enhancements based on the Hackathon in January 2025 by@wedeling in#440
  • Changed sc_analysis.py to suppress divide by zero in Sobol index calculation, if a zero is present. (Unhandled divide by zero case in sc_analysis.py#415 )
  • fixed plotting error of analysis.analysis.adaptation_table() fails#430
  • Failures when running the non-regression tests manually (1)#387 closed
• Bump jinja2 from 3.1.5 to 3.1.6 by@dependabot in#443
• Improves EasyVVUQ installation script with better user experience by@mzrghorbani in#442
• Fixed pip installation (issue#445 by@wedeling in#446 )

New Contributors

• @dependabot made their first contribution in#443

Full Changelog:v1.2.2.3...v1.3

Contributors

Test for PyPi.

A few minor but important updates in this release:

• Fixed a wide range of tests, library and dependency issues.
• Incorporated a range of documentation improvements.
• Added an example for using PCE with aleatoric uncertainty

Note that the readthedocs page is likely to be further expanded and updated in the weeks after this release.

This is the July 2023 release of EasyVVUQ, as part of the SEAVEAtk, with the following minor updates:

Fixes and updates

• Fixed several tests for newer Python versions.
• Updated integration with QCG-PilotJob

Tutorials

• SSC tutorial:https://github.com/UCL-CCS/EasyVVUQ/blob/dev/tutorials/simplex_stochastic_collocation_tutorial.ipynb
• Hyperparameter tuning tutorial, local sampling:https://github.com/UCL-CCS/EasyVVUQ/blob/dev/tutorials/hyperparameter_tuning_tutorial.ipynb
• Hyperparameter tuning tutorial, remote sampling with FabSim3:https://github.com/UCL-CCS/EasyVVUQ/blob/dev/tutorials/hyperparameter_tuning_tutorial_with_fabsim.ipynb

This is the March 2023 release of EasyVVUQ, as part of the SEAVEAtk, with the following updates:

New features

• New Simplex Stochastic Collocation sampler for irregular outputs, e.g. with discontinuities or high gradients in the stochastic input space. Works for scalar QoI only thus far.
• Grid-Search sampler, (e.g. for neural-network hyper parameter tuning).
• HDF5 decoder to allow for reading HDF5 output files, useful when dealing with outputs of different size.

Tutorials

• SSC tutorial:https://github.com/UCL-CCS/EasyVVUQ/blob/dev/tutorials/simplex_stochastic_collocation_tutorial.ipynb
• Hyperparameter tuning tutorial, local sampling:https://github.com/UCL-CCS/EasyVVUQ/blob/dev/tutorials/hyperparameter_tuning_tutorial.ipynb
• Hyperparameter tuning tutorial, remote sampling with FabSim3:https://github.com/UCL-CCS/EasyVVUQ/blob/dev/tutorials/hyperparameter_tuning_tutorial_with_fabsim.ipynb

Usability updates

• Make it more obvious how to import a pandas dataframe containing cases to be considered
• Make it more obvious how to massage the results from the runs before performing the PCE/SC/MC analysis

Overhaul of SC sampler / analysis class:

• Made isotropic sparse-grid subroutines more scalable to higher input dimensions. Reused dimension-adaptive subroutines for this purpose, instead of having (slower) separate isotropic routines.
• Rewrote dimension-adaptive SC expansion as a standard PCE expansion with generalized PCE coefficients. See adaptive sparse-grid tutorial.

Documentation:

• Extensive methodological sparse-grid tutorial:https://www.researchgate.net/publication/359296270_Adaptive_sparse-grid_tutorial
• New tutorial on using mathematical expressions involving parameters in template files using the Jinja encoder:https://github.com/UCL-CCS/EasyVVUQ/blob/dev/tutorials/jinja_tutorial.ipynb

New features:

• Updated the documentation in a range of places.

Bug fixes:

*Fixed direct integration of EasyVVUQ with QCG-PilotJob. Previously there was an issue with large campaigns where the integration could fail due to an excessively long command-line argument.
*Fixed bug where unsuitable models could be applied with QCG-PilotJob integration.
*Fixed MC sampler for use with 1 parameter:fac0b57

Tutorials:

• Added an example for including noise in an EasyVVUQ campaign ( easyvvuq_Ishigami_with_noise_tutorial.ipynb)

New features:

• Ability to add external runs via a DataFrame
• Ability to execute EasyVVUQ workflows from R

Tutorials:

• Updates to Dimension Adaptive Fusion tutorial.

EasyVVUQ v1.0

New Features

• Better support to execute pure Python simulations.
• Added a surrogate method to AnalysisResults classes.
• QCG-PJ support.
• Gaussian Process Surrogate analysis method.
• Reworked Campaign and hand optimised database.
• Re-implemented Actions system for modular execution options.
• DataFrameSampler for uploading new

Updates

• Large scale code refactoring.
• Docstring and documentation updates.
• Additional testing and benchmarking.
• Continuous benchmarking.

A bug-fix release.