v1.1.0

⚠️ Breaking changes ⚠️

• Changed behaviour of Simulator.forward_batch (returns both y and x)
• AutoEmulate no longer takes in model_tuning argument, model tuning is not done if model_params != None, to use default params without tuning use model_params = {}

What's Changed

• Clear nb outputs by @radka-j in #685
• Update HMW notebook by @radka-j in #687
• Undo changes after experimental merge by @radka-j in #689
• Update tutorial title by @radka-j in #699
• Fix math in docstring by @radka-j in #703
• Fix install link in README by @radka-j in #705
• add reaction-diffusion gif by @ContiPaolo in #706
• return dict of results from simulator by @marjanfamili in #710
• Speed up tests by @EdwinB12 in #686
• remove poetry support by @EdwinB12 in #700
• Update contributing docs by @radka-j in #715
• Add emulator info by @radka-j in #701
• Pre-commit hook to strip nb output by @radka-j in #713
• Fix nbstripout issue by @radka-j in #717
• 681 add testing support for different os by @EdwinB12 in #702
• Use compare method in dim reduction tutorial by @radka-j in #714
• Add load_model method and update quickstart by @sgreenbury in #729
• Move logo to package, update pyproject.toml by @sgreenbury in #731
• Add AutoEmulate class to init to allow import from top level of package. by @EdwinB12 in #724
• Add option to save plots to file by @radka-j in #734
• Fix package data and revert logo path by @sgreenbury in #739
• Fix order of sobol indexes when creating df by @radka-j in #738
• Update install instructions by @radka-j in #736
• Revise transforms by @sgreenbury in #726
• Add support for optional bootstrap by @sgreenbury in #742
• Add transformed emulator params by @sgreenbury in #744
• Restructure tests by @radka-j in #747
• Add option to subset models by @radka-j in #753
• Re-initialise emulator before refitting in HMW by @radka-j in #752
• Add cloud sampling to HM by @radka-j in #725
• Add API to support plotting subset of data by @sgreenbury in #755
• Move import outside of test function by @EdwinB12 in #762
• Add API to support fixing GP mean and kernel hyperparameters during training by @sgreenbury in #761
• Add QMC Sobol sampling by @radka-j in #754
• Fix assertion and class resolution by @sgreenbury in #767
• Add experimental dataset, model and notebooks for spatio-temporal problems by @sgreenbury in #770
• Update BayesianCalibationto optionally include model variance and discrepancy by @sgreenbury in #765
• Add GP subclasses and functionality to generate new subclasses easily by @sgreenbury in #764
• Set random seeds in Bayes tutorial for reproducibility by @radka-j in #768
• should report failed doc builds by @EdwinB12 in #766
• Fix simulator setter bug by @radka-j in #777
• Ruff ignore arguments by @marjanfamili in #776
• add random seed into bayesian tutorial by @EdwinB12 in #779
• Build docs with autoemulate installed from branch by @radka-j in #781
• Update Simulator.forward_batch to always return x,y by @radka-j in #772
• Update calibration tutorial PR suggestions by @sgreenbury in #791
• Update calibration tutorial by @radka-j in #782
• Don't tune if user passes model_params by @radka-j in #780
• Fix tutorial by @radka-j in #793
• Refactor Naghavi case study by @radka-j in #632
• added README to case_studies directory by @ritkaarsingh30 in #799
• Add spatiotemporal deps as optional extra by @sgreenbury in #792
• Set log level in HM tutorials by @radka-j in #802
• Update version by @radka-j in #798

New Contributors

• @ritkaarsingh30 made their first contribution in #799

Full Changelog: v1.0.0...v1.1.0

v1.0.0

What's Changed

• Add MLP by @sgreenbury in #559
• Handle random_seed setting at higher level in package by @edwardchalstrey1 in #512
• Docs restructure by @radka-j in #573
• Refactor: add transforms and transformed emulator by @sgreenbury in #474
• Add Polynomials to experimental by @edwardchalstrey1 in #568
• Add more emulator subclasses by @sgreenbury in #561
• Fix MLP initialization by @sgreenbury in #581
• 492 simulator code contribution tutorial by @EdwinB12 in #577
• Add Radial Basis functions to experimental & update Tuner run loop by @edwardchalstrey1 in #567
• Add EarlyStopping for GPs by @radka-j in #587
• docs: add Apple-silicon test note (Refs #410) by @nayara-focs in #590
• changed the numbering of tutorial 08 to 07 by @marjanfamili in #592
• Add ensembles by @cisprague in #560
• Add GP with correlated tasks by @sgreenbury in #585
• Hmc re refactor by @marjanfamili in #599
• Add support for transforms of MultivariateNormal by @sgreenbury in #586
• Fix training only first batch by @sgreenbury in #604
• Add transforms to compare loop by @sgreenbury in #551
• Add learning rate scheduler by @edwardchalstrey1 in #582
• Grad handling by @radka-j in #605
• Refactor HMC calibration by @radka-j in #575
• 428 improve logging by @EdwinB12 in #608
• Set RBF supports_grad=False by @radka-j in #616
• Evaluate on test data, update compare loop and add Results/plots by @edwardchalstrey1 in #603
• Start experimental docs by @radka-j in #598
• Add bootstrap, fix evaluation and revise results by @sgreenbury in #620
• Add logging to Active Learner by @edwardchalstrey1 in #624
• Update docstrings by @radka-j in #626
• Common scheduler setup function for all emulators by @edwardchalstrey1 in #623
• Fix missing model config to pass to cv by @sgreenbury in #627
• Add logger to BayesianCalibrator by @edwardchalstrey1 in #628
• Remove experimental (v1) code reliance on v0 codebase by @radka-j in #630
• Fix reaction-diffusion example and update GP predictive by @sgreenbury in #639
• Revise GP predict approach by @sgreenbury in #638
• History Matching tutorials by @radka-j in #625
• Bayesian calibration tutorial by @radka-j in #636
• Add str and dict serialization option for emulators and transforms in AutoEmulate class by @sgreenbury in #641
• Add saving and loading models to experimental by @edwardchalstrey1 in #629
• Add fixes from further testing and benchmarking by @sgreenbury in #647
• Experimental sensitivity analysis tutorial by @edwardchalstrey1 in #642
• Add handling for failed simulations by @radka-j in #643
• Add API to update simulator parameters and enable sampling when equal bounds by @sgreenbury in #660
• Add standardization for x and y to Emulator by @sgreenbury in #650
• Remove obsolete preprocessing by @sgreenbury in #666
• Add docs lint config and update docstrings by @sgreenbury in #663
• Add script for benchmarking simulators with different parameters by @sgreenbury in #621
• Revise config to params by @sgreenbury in #667
• Update module structure to add core module by @sgreenbury in #670
• Fix pre-commit by @sgreenbury in #669
• Update config for experimental and API docs by @sgreenbury in #664
• Move modules from experimental, move current to v0 by @sgreenbury in #672
• Omit v0/ from coverage by @sgreenbury in #673
• Fix docs notebooks by @sgreenbury in #674
• Add datasets to package and update path by @sgreenbury in #675
• Update default to enable tuning by @sgreenbury in #677
• Update version to 1.0.0 by @sgreenbury in #678

New Contributors

• @nayara-focs made their first contribution in #590

Full Changelog: v0.4.0...v1.0.0

v0.4.0

Major Changes

• rm CNPs by @radka-j in #554

Other Changes

• Update quickstart tutorial by @radka-j in #523
• Update Simulator base class by @radka-j in #520
• Post-workshop improved tutorial setup instructions by @edwardchalstrey1 in #535
• Add target-version to ruff to extend lints by @sgreenbury in #541
• Refactor sensitivity analysis by @radka-j in #539
• Bug fixes by @radka-j in #556
• Add test for device to history matching by @sgreenbury in #562
• HM refactor by @radka-j in #504
• Add Gradient boosting to experimental by @edwardchalstrey1 in #565
• Execute nb when building docs except workflow example by @radka-j in #569
• Bump version to 0.4 by @radka-j in #570

Full Changelog: v0.3.3...v0.4.0

v0.3.3

What's Changed

• Add epidemic and projectile Simulator by @radka-j in #516
• Mcmc hmc by @marjanfamili in #505
• Update simulators to return arrays of right shape by @radka-j in #518

Full Changelog: v0.3.2...v0.3.3

v0.3.2

What's Changed

• Set random seeds for Emulator models by @edwardchalstrey1 in #479
• Update dashboard to clear previous plot by @radka-j in #510
• Small bug fix in HM by @marjanfamili in #507

Full Changelog: v0.3.1...v0.3.2

v0.3.1

What's Changed

• fix triple plot in dashboard by @marjanfamili in #442
• Fix tutorial rendering on docs page by @radka-j in #443
• Update tutorial by @radka-j in #445
• Update docs to not execute workflow notebook by @radka-j in #446
• Revise base emulator and update emulators by @sgreenbury in #453
• 400 revise input types by @EdwinB12 in #456
• Refactor sklearn models for experimental by @edwardchalstrey1 in #415
• Add exploratory notebook for the well active matter dataset by @edwardchalstrey1 in #476
• Update HistoryMatching by @radka-j in #478
• Refactor/422 - consistent Utils for experimental code by @edwardchalstrey1 in #464
• Add device handling by @sgreenbury in #458
• Add checks for outputs in ValidationMixin by @edwardchalstrey1 in #489
• Ensure using get_model with a named model works as expected by @edwardchalstrey1 in #496
• Morris Sensitivity Analysis - update workflow by @marjanfamili in #493
• Add option to disable progress bar by @sgreenbury in #448
• Bump version to 0.3.1 by @edwardchalstrey1 in #502

Full Changelog: v0.3.0...v0.3.1

v0.3.0

New features

• History matching (#294)
• Dimensionality reduction (#390)
• Active learning (experimental) (#358)
• End-to-end workflow example with a user-provided simulator (#423)
• Initial PyTorch refactor (experimental)

What's Changed

• Release/v0.2.1 by @mastoffel in #289
• Docs fixes by @mastoffel in #290
• Community by @mastoffel in #296
• AutoEmulate website by @mastoffel in #303
• PyTorch model extractor by @mastoffel in #298
• Website by @mastoffel in #309
• added history matching and test by @marjanfamili in #297
• Use SALib ResultDict.to_df() function to provide results in pandas dataframe format by @willu47 in #305
• Show plots by @mastoffel in #310
• Sensitivity analysis var name by @marjanfamili in #300
• Minor docs update by @radka-j in #333
• Refactor documentation site to include new logo and link to new website by @edwardchalstrey1 in #338
• Cardiac tutorial by @aranas in #233
• Update config (#360) by @sgreenbury in #365
• 322 base class design by @EdwinB12 in #360
• Preprocessing pipeline by @ContiPaolo in #363
• Add Tuner by @radka-j in #380
• Extend InputTypeMixin by @radka-j in #382
• Add typing by @sgreenbury in #383
• Revise tuner and base class by @sgreenbury in #387
• Cross validate with torchmetrics by @sgreenbury in #392
• Add cross_validate function by @radka-j in #388
• Make docs only run on pushes and PR to main by @EdwinB12 in #395
• Update pyright config by @sgreenbury in #389
• Tutorial and getting started updates by @edwardchalstrey1 in #396
• Update ruff config by @sgreenbury in #391
• 301 add modular active learning by @cisprague in #358
• Add Gaussian Process for refactor by @sgreenbury in #384
• Several easy docs fixes by @edwardchalstrey1 in #401
• abstract base simulator by @marjanfamili in #409
• Preprocessing by @marjanfamili in #390
• Add refactored LGBM model to experimental emulators by @edwardchalstrey1 in #399
• Adding Conditonal Neural Network by @EdwinB12 in #398
• Add compare for experimental by @sgreenbury in #397
• Fix error raised in sensitivity analysis docs by @sgreenbury in #431
• Added stream-based active learning schematic. by @cisprague in #433
• Refactor active learning by @sgreenbury in #411
• History matching refine model by @marjanfamili in #423
• Add workflow tutorial to docs by @radka-j in #440
• Update version to 0.3.0 by @sgreenbury in #439

New Contributors

• @marjanfamili made their first contribution in #297
• @willu47 made their first contribution in #305
• @radka-j made their first contribution in #333
• @edwardchalstrey1 made their first contribution in #338
• @sgreenbury made their first contribution in #365
• @EdwinB12 made their first contribution in #360
• @ContiPaolo made their first contribution in #363
• @cisprague made their first contribution in #358

Full Changelog: v0.2.1...v0.3.0

v0.2.1

• fixed sci-kit learn dependency bug during installation
• fixed warnings in tests

v0.2.0

see CHANGELOG for details.

v0.1.0.post1

Pre-release with basic functionality to automatically build emulators from simulation data, including:

• 10 machine learning models commonly used in surrogate modeling, incl. neural nets
• all models are scikit-learn compatible
• cross-validation
• hyperparameter optimisation (random and bayesian)
• parallelisation
• data pre-processing: scaling, dimensionality reduction
• plotting