Change log¶
All notable changes to this project will be documented in this file.
The format is based on 'Keep a Changelog' and this project adheres to semantic versioning.
[2.0.0] — 2020-11-10¶
Added¶
Updating moment estimators in small batches (see
MonteCarloIndex.eventGroupSize
) (MR !15).Asynchronous algorithms for single- and multi-level Monte Carlo methods (MR !16)
Estimation of 'combined' expectation (time and events) for random processes (MR !19)
Support for multi-valued (e.g. vectors) random variables (issue #31, MR !36)
Estimation of moments of order 3 and 4, with a posteriori estimation of statistical error (MR !36)
Examples, both built-in and using Kratos as external solver (MR !38)
Test cases
README (MR !33)
Changelog (MR !42)
New features of the Kratos interface:
different meshing strategies (support distributed environments)
write to file in distributed environments
Changed¶
Documentation re-created anew, from docstrings (MR !31, !32, !41)
The choice of framework for distributed computing is contained in a single module definition (issue #30, MR !30)
Fixed¶
Excessive memory usage in case of high number of estimators (MR !23).
Deprecated¶
∅
Removed¶
∅
[1.0.0] — 2019-05-31¶
Initial release.
Added¶
Single- and multi-level Monte Carlo estimation of expectation and variance of real-valued random variable
Adaptivity of number of samples and levels based on a posteriori error estimators
Stopping criterion based on a posteriori error estimators
Interface with external solver: Kratos
Parallelisation using the Python interface common to COMPSs and HyperLoom