Skip to content
Snippets Groups Projects
Commit 50cd35d4 authored by kohlhaasrebecca's avatar kohlhaasrebecca
Browse files

Started Changelog

parent 038ef383
No related branches found
No related tags found
2 merge requests!22Update bvr in branch,!17Version 1.0.0
# CHANGELOG
## [Unreleased]
### Requirements
* numpy now at 1.23.3
* ....
### Added
* PyLinkForwardModel has new link_type 'umbridge' for UM-Bridge type models
* class `InputSpace` as parent class to `ExpDesigns`
* `MetaModel` has option to train with OLs+constraints, this is toggled by `MetaModel.apply_constraints`
* Example `umbridge_tsunamitutorial` to show two options of using UM-Bridge type models
* Example `convergence_tests` for the constraints
### Changed
* `MetaModelEngine` renamed to `Engine`
* Function `hellinger_distance()` moved out of `MetaModelEngine`
* Function `logpdf()` moved out of `MetaModelEngine`
* Function `subdomain()` moved out of `MetaModelEngine`
* Renamed `MetaModelEngine.run` and split into `Engine.train_normal` and `Engine.train_sequential`
* `MetaModelEngine.opt_SeqDesign()` renamed to `Engine.choose_next_sample()`
* `Engine.eval_metamodel()` equivalent to previous `MetaModel.eval_metamodel()`
* `MetaModel` constructor does not receive the model
* `MetaModel` fully independent of model
* `MetaModel.add_ExpDesign()` repurposed for the new `InputSpace` class and renamed to `MetaModel.addInputSpace()`
* `MetaModel.create_metamodel()` split into `MetaModel.build_metamodel()` and `MetaModel.fit()`
* `MetaModel.fit()` renamed to `MetaModel.regression()`
* `MetaModel.eval_metamodel()` only runs on given samples, option to give number of wanted evaluations removed
* Function `compute_pce_moments()` moved from `MetaModelEngine` to `MetaModel`
* Function `generate_polynomials()` moved from `ExpDesigns` to `MetaModel`
* Read-in of MC-references performed by `PyLinkForwardModel.read_observation()`, formerly in function `PyLinkForwardModel.read_mc_reference()`
* `ExpDesigns` now independent of `MetaModel`
* Parameters related to sequential training were moved from `MetaModel` to `ExpDesigns` (`Seq*`, `valid_model_runs`)
* Use `ExpDesigns.Y` instead of `MetaModel.ModelOutputDict`
* Constructors of `PostProcessing`, `BayesInference`, `BayesModelComp` are to be given the engine instead of the metamodel
* Constructor of `Exploration` to be given `ExpDesigns` instead of `MetaModel`
* `within_change()` moved out of `PyLinkForwardModel`
* Examples updated to these changes
### Removed
* Class `SequentialDesign`
* `ExpDesigns.method`
* `MetaModel.create_basis_indices()`
......@@ -128,7 +128,6 @@ if __name__ == "__main__":
ExpDesign = ExpDesigns(Inputs)
# One-shot (normal) or Sequential Adaptive (sequential) Design
ExpDesign.method = 'normal'
ExpDesign.n_init_samples = 100
# Sampling methods
......
......@@ -116,7 +116,6 @@ if __name__ == "__main__":
ExpDesign = ExpDesigns(Inputs)
# One-shot (normal) or Sequential Adaptive (sequential) Design
ExpDesign.method = 'normal'
ExpDesign.n_init_samples = 50
# Sampling methods
......
......@@ -178,7 +178,6 @@ if __name__ == "__main__":
L2_ExpDesign = ExpDesigns(L2_Inputs)
# One-shot (normal) or Sequential Adaptive (sequential) Design
L2_ExpDesign.method = 'normal'
L2_ExpDesign.n_init_samples = 100
# Sampling methods
......
This diff is collapsed.
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment