CHANGELOG.md

Unreleased

Features

• #918 - Adds a plot for predictions sampled from a posterior distribution (pybop.plot.predictive).
• #940 - Adds support for Python 3.14 (EP-BOLFI optimiser and PyProBE still restricted to Python 3.12 or below).

Optimisations

• #925 - Add UnboundedDistribution and the get_transformed_distribution functionality.

Bug Fixes

• #915 - Fixes axis labels for non-standard domain names, adds Dataset length property and adds kind property to Interpolant.
• #911 - Fixes the passing of the cost log to the Voronoi surface plot.
• #905 - Remove restriction on numpy.

Breaking Changes

• #942 - Adds evaluate_batch to the costs and ensures that an Evaluation is returned.

v26.3 - 2026-03-05

Features

• #897 - Adds separate LogPrior, LogPDF and LogPosterior classes and updates set_target.
• #873 - Adds methods for saving result and reconstructing result from saved data. result.save: saves entire python object using pickle. result.save_data: saves primarily the logger data and any other data required to reconstruct the result from the problem or the sampler (for SamplingResult). Result.load_result: reconstructs the Result object based on the underlying problem (or sampler for SamplingResult) and the data saved to file.
• #862 - Adds pybop.MarginalDistribution, pybop.MultivariateLogNormal.
• #889 - Adds methods for setting the initial state from a voltage to the grouped models.
• #869 - Adds methods for pre-processing current data for linear interpolation.
• #868 - Adds support for Python3.13 (NumPy restricted to <2.4, EP-BOLFI optimiser and PyProBE do not support Python 3.13).
• #871 - Adds a lumped thermal model called CellTemperature.
• #846 - Adds Bayesian optimisation framework and, as an example, the EP-BOLFI optimiser

Optimisations

Bug Fixes

• #890 - Fix the assignment of parameters within a MetaProblem.
• #847 - Update readme and diagram of pybop components so that the diagram is displayed correctly in the readme.

Breaking Changes

• #894 - Distinguish different uses of sigma, pass the covariance to the samplers, and add parameter get_mean and get_std functions.
• #862 - Removes MultivariateParameters class. Instead allows multivariate parameters to be passed via pybamm.ParameterValues (as a pybop.Parameter with a pybop.MarginalDistribution). The pybop.Parameters class now handles multivariate parameters. Multivariate distributions are now defined in the model space instead of the search space.
• #878 - Use "Current [A]" instead of "Current function [A]" in datasets and allow list of control functions.
• #864 - Remove check_already_exists from ParameterValues following PyBaMM PR 5339.
• #860 - Create a parent class for optimisation and sampling results, move PosteriorSummary attributes to the SamplingResult and deprecate the pints.AdaptiveCovarianceMCMC sampler.
• #857 - Deprecate the custom PyBaMM model build process for a simulation without an experiment and rename batch_solve as solve_batch to align with other functions.
• #839 - Renames 'prior' as 'distribution' for pybop.Parameter. Allows construction of a pybop.Parameter with a distribution of type scipy.stats.distributions.rv_frozen. Removes margins, set_bounds, remove_bounds from pybop.Parameter.

v25.11 - 2025-11-24

Features

• #815 - Adds function import_pyprobe_result to import a pyprobe.result into a pybop.dataset. Allows for creating a dataset directly from a pybamm.solution object.

Optimisations

• #837 - Update the descriptons in the example scripts and notebooks.
• #833 - Upgrade to Pints 0.5.1, PyBaMM 25.10.1 and NumPy 2, fix some deprecation warnings.
• #816 - Enable simulator multi-processing via the evaluators.

Bug Fixes

• #834 - Finite difference calculations of the Hessian matrix are updated. A new notebbok file is added which demonstrates sensitivity analysis using SALib.

Breaking Changes

• #829 - Create SamplingResult and best inputs property for results.

v25.10 - 2025-10-31

This release presents a major restructure of PyBOP's base classes. We move from setting up a model, problem, cost, then optimiser to defining a simulator, cost, problem, and then optimiser. A pybop.pybamm.Simulator is designed to simulate a pybamm.BaseModel. Optimisation parameters can be passed through a pybamm.ParameterValues class. To understand how to update your use of PyBOP, please take a look at the example notebooks and scripts.

Breaking Changes

• #820 - Remove the name property from pybop.Parameter
• #821 - Remove the papers folder and update Readme.
• #809 - Major restructure, including:
• Deprecate Python 3.9 support
• Update initial state setting (requires PyBaMM > 25.8)
• Remove jax methods
• Add PyBaMM and PyBaMM-EIS simulators for rebuilding and running simulations for a given set of input parameters
• Remove PyBaMM wrappers and enable use of PyBaMM model, parameter values and experiment classes
• Remove observers
• Remove standalone class examples
• Improve logging
• Remove Optimisation and MCMCSampler wrapper classes
• Remove Fisher information computation
• Rename apply_transform argument to transformed
• Remove the update_capacity option from the DesignProblem
• Update sensitivities retrieval (for PyBaMM 25.8)
• Remove uninformative examples
• Move optimiser and sampler options into defined classes
• Add PyBaMM utilities, design variable definitions and the add_variable_to_model function
• Allow plotting via functions on the OptimisationResult
• Separate the cost classes from the Problem
• Replace FittingProblem and DesignProblem by a single Problem class
• Rename and reimplement MultiFittingProblem as MetaProblem
• Add BaseSimulator as a generic base class for the pybop.pybamm.Simulator and pybop.pybamm.EISSimulator
• Enable pybop.Parameter objects to be passed directly to an instance of pybamm.ParameterValues
• Update the method for setting formation concentrations to be part of the model definition
• Rename some example scripts and notebooks
• Update the docs and test workflows

v25.6 - 2025-07-16

Features

• #767 - Adds the GroupedSPM model for parameter identification.
• #644 - Adds example applications for common battery experiments.

Optimisations

• #763 - Updates the GITT pulse fitting method to allow iteration over many pulses.

Bug Fixes

• #771 - Match naming of n_sensitivity_samples and fix intermittent test_optimisation_f_guessed test.
• #737 - Sensitivities no longer available for CasadiSolver in Pybamm v25.6 onwards. Updates Hallemans example scripts.
• #705 - Bug fix fitting_problem.evaulate() failure return type alongside fixes for Pybamm v25.4.
• #546 - Default Pybamm solver to IDAKLU, changes required for Pybamm v25.4.1

Breaking Changes

v25.3 - 2025-03-28

Features

• #649 - Adds verbose outputs to Pints-based optimisers.
• #659 - Enables user-defined weightings of the error measures.
• #674 - Adds the reason for stopping to the OptimisationResult.
• #663 - Adds DFN fitting examples alongside synthetic data generation methods.

Optimisations

• #676 - Update the format of the problem sensitivities to a dict.
• #681 - Update the spatial variable defaults of the GroupedSPMe model.
• #692 - Improvements/fixes for BaseSampler and BasePintsSampler classes, adds ChainProcessor classes w/ clearer structure.

Bug Fixes

• #678 - Fixed bug where model wasn't plotted for observer classes with pybop.plot.quick().

Breaking Changes

• #684 - Updates plot.quick to plot.problem for clarity.
• #661 - Adds pybop.CostInterface which aligns the optimisers and samplers with a unified call_cost in which transformations and sign inversions are applied. Also includes bug fixes for transformations and gradient calculations.

v25.1 - 2025-02-03

Features

• #636 - Adds pybop.IRPropPlus optimiser with corresponding tests.
• #635 - Adds support for multi-proposal evaluation of list-like objects to BaseCost classes.
• #635 - Adds global parameter sensitivity analysis with method BaseCost.sensitivity_analysis. This is computation is added to OptimisationResult if optimiser arg compute_sensitivities is True. An additional arg is added to select the number of samples for analysis: n_sensitivity_samples.
• [#630] (#632) - Fisher Information Matrix added to BaseLikelihood class.
• #619 - Adds pybop.SimulatingAnnealing optimiser with corresponding tests.
• #565 - DigiBatt added as funding partner.

Optimisations

• #638 - Allows the problem class to accept any domain name.
• #618 - Adds Mean Absolute Error (MAE) and Mean Squared Error (MSE) costs.
• #601 - Deprecates MultiOptimisationResult by merging with OptimisationResult.
• #600 - Removes repetitious functionality within the cost classes.

Bug Fixes

• #602 - Aligns the standard quick plot of MultiFittingProblem outputs.

Breaking Changes

• #656 - Completes ParameterSet changes from #593 and aligns the simulation options in model.predict with the model properties such as the solver.
• #593 - Enables ParameterSet to systematically return a pybamm.ParameterValues object within the model class.

v24.12 - 2024-12-21

Features

• #481 - Adds experimental support for PyBaMM's jaxified IDAKLU solver. Includes Jax-specific cost functions pybop.JaxSumSquareError and pybop.JaxLogNormalLikelihood. Adds Jax optional dependency to PyBaMM dependency.
• #597 - Adds number of function evaluations n_evaluations to OptimisationResult.
• #362 - Adds the classify_using_Hessian functionality to classify the optimised result.
• #584 - Adds the GroupedSPMe model for parameter identification.
• #571 - Adds Multistart functionality to optimisers via initialisation arg multistart.
• #582 - Fixes population_size arg for Pints' based optimisers, reshapes parameters.rvs to be parameter instances.
• #570 - Updates the contour and surface plots, adds mixed chain effective sample size computation, x0 to optim.log
• #566 - Adds UnitHyperCube transformation class, fixes incorrect application of gradient transformation.
• #569 - Adds parameter specific learning rate functionality to GradientDescent optimiser.
• #282 - Restructures the examples directory.
• #396 - Adds ecm_with_tau.py example script.
• #452 - Extends cell_mass and approximate_capacity for half-cell models.
• #544 - Allows iterative plotting using StandardPlot.
• #541 - Adds ScaledLogLikelihood and BaseMetaLikelihood classes.
• #409 - Adds plotting and convergence methods for Monte Carlo sampling. Includes open-access Tesla 4680 dataset for Bayesian inference example. Fixes transformations for sampling.
• #531 - Adds Voronoi optimiser surface plot (pybop.plot.surface) for fast optimiser aligned cost visualisation.
• #532 - Adds linked_parameters example script which shows how to update linked parameters during design optimisation.
• #529 - Adds GravimetricPowerDensity and VolumetricPowerDensity costs, along with the mathjax extension for Sphinx.

Optimisations

• #580 - Random Search optimiser is implimented.
• #588 - Makes minimising a property of BaseOptimiser set by the cost class.
• #512 - Refactors LogPosterior with attributes pointing to composed likelihood object.
• #551 - Refactors Optimiser arguments, population_size and max_iterations as default args, improves optimiser docstrings

Bug Fixes

• #595 - Fixes non-finite LogTransformed bounds for indices of zero.
• #561 - Bug fixes the sign of the SciPy cost logs for maximised costs.
• #505 - Bug fixes for LogPosterior with transformed GaussianLogLikelihood likelihood.

Breaking Changes

• #481 - problem.model is now a copied instance of model
• #598 - Depreciated Adam optimiser has been removed, see AdamW for replacement.
• #531 - Plot methods moved to pybop.plot with mostly minimal renaming. For example, pybop.plot_parameters is now pybop.plot.parameters. Other breaking changes include: pybop.plot2d to pybop.plot.contour.
• #526 - Refactor OptimisationResults classes, with optim.run() now return the full object. Adds finite cost value check for optimised parameters.

v24.9.1 - 2024-09-16

Features

Bug Fixes

• #495 - Bugfixes for Transformation class, adds apply_transform optional arg to BaseCost for transformation functionality.

Breaking Changes

v24.9.0 - 2024-09-10

Features

• #462 - Enables multidimensional learning rate for pybop.AdamW with updated (more robust) integration testing. Fixes bug in Minkowski and SumofPower cost functions for gradient-based optimisers.
• #411 - Updates notebooks with README in examples/ directory, removes kaleido dependency and moves to nbviewer rendering, displays notebook figures with notebook_connected plotly renderer
• #6 - Adds Monte Carlo functionality, with methods based on Pints' algorithms. A base class is added BaseSampler, in addition to PintsBaseSampler.
• #353 - Allow user-defined check_params functions to enforce nonlinear constraints, and enable SciPy constrained optimisation methods
• #222 - Adds an example for performing and electrode balancing.
• #441 - Adds an example for estimating constants within a pybamm.FunctionalParameter.
• #405 - Adds frequency-domain based EIS prediction methods via model.simulateEIS and updates to problem.evaluate with examples and tests.
• #460 - Notebook example files added for ECM and folder structure updated.
• #450 - Adds support for IDAKLU with output variables, and corresponding examples, tests.
• #364 - Adds the MultiFittingProblem class and the multi_fitting example script.
• #444 - Merge BaseModel build() and rebuild() functionality.
• #435 - Adds SLF001 linting for private members.
• #418 - Wraps the get_parameter_info method from PyBaMM to get a dictionary of parameter names and types.
• #413 - Adds DesignCost functionality to WeightedCost class with additional tests.
• #357 - Adds Transformation() class with LogTransformation(), IdentityTransformation(), and ScaledTransformation(), ComposedTransformation() implementations with corresponding examples and tests.
• #427 - Adds the nbstripout pre-commit hook to remove unnecessary metadata from notebooks.
• #327 - Adds the WeightedCost subclass, defines when to evaluate a problem and adds the spm_weighted_cost example script.
• #393 - Adds Minkowski and SumofPower cost classes, with an example and corresponding tests.
• #403 - Adds lychee link checking action.

Bug Fixes

• #473 - Bugfixes for transformation class, adds optional apply_transform arg to BaseCost.__call__(), adds log_update() method to BaseOptimiser.
• #464 - Fix order of design parameter_set updates and refactor update_capacity.
• #468 - Renames quick_plot.py to standard_plots.py.
• #454 - Fixes benchmarking suite.
• #421 - Adds a default value for the initial SOC for design problems.

Breaking Changes

• #499 - BPX is added as an optional dependency.
• #483 - Replaces pybop.MAP with pybop.LogPosterior with an updated call args and bugfixes.
• #436 - API Change: The functionality from BaseCost.evaluate/S1 & BaseCost._evaluate/S1 is represented in BaseCost.__call__ & BaseCost.compute. BaseCost.compute directly acts on the predictions, while BaseCost.__call__ calls BaseProblem.evaluate/S1 before BaseCost.compute. compute has optional args for gradient cost calculations.
• #424 - Replaces the init_soc input to FittingProblem with the option to pass an initial OCV value, updates BaseModel and fixes multi_model_identification.ipynb and spm_electrode_design.ipynb.

v24.6.1 - 2024-07-31

Features

• #313 - Fixes for PyBaMM v24.5, drops support for PyBaMM v23.9, v24.1

Bug Fixes

Breaking Changes

v24.6 - 2024-07-08

Features

• #319 - Adds CuckooSearch optimiser with corresponding tests.
• #359 - Aligning Inputs between problem, observer and model.
• #379 - Adds model.simulateS1 to weekly benchmarks.
• #174 - Adds new logo and updates Readme for accessibility.
• #316 - Adds Adam with weight decay (AdamW) optimiser, adds depreciation warning for pints.Adam implementation.
• #271 - Aligns the output of the optimisers via a generalisation of Result class.
• #315 - Updates init structure to remove circular import issues and minimises dependancy imports across codebase for faster PyBOP module import. Adds type-hints to BaseModel and refactors rebuild parameter variables.
• #236 - Restructures the optimiser classes, adds a new optimisation API through direct construction and keyword arguments, and fixes the setting of max_iterations, and _minimising. Introduces pybop.BaseOptimiser, pybop.BasePintsOptimiser, and pybop.BaseSciPyOptimiser classes.
• #322 - Add Parameters class to store and access multiple parameters in one object.
• #321 - Updates Prior classes with BaseClass, adds a problem.sample_initial_conditions method to improve stability of SciPy.Minimize optimiser.
• #249 - Add WeppnerHuggins model and GITT example.
• #304 - Decreases the testing suite completion time.
• #301 - Updates default echem solver to "fast with events" mode.
• #251 - Increment PyBaMM > v23.5, remove redundant tests within integration tests, increment citation version, fix examples with incorrect model definitions.
• #285 - Drop support for Python 3.8.
• #275 - Adds Maximum a Posteriori (MAP) cost function with corresponding tests.
• #273 - Adds notebooks to nox examples session and updates CI workflows for change.
• #250 - Adds DFN, MPM, MSMR models and moves multiple construction variables to BaseEChem. Adds exception catch on simulate & simulateS1.
• #241 - Adds experimental circuit model fitting notebook with LG M50 data.
• #268 - Fixes the GitHub Release artifact uploads, allowing verification of codesigned binaries and source distributions via sigstore-python.
• #79 - Adds BPX as a dependency and imports BPX support from PyBaMM.
• #267 - Add classifiers to pyproject.toml, update project.urls.
• #195 - Adds the Nelder-Mead optimiser from PINTS as another option.

Bug Fixes

• #393 - General integration test fixes. Adds UserWarning when using Plot2d with prior generated bounds.
• #338 - Fixes GaussianLogLikelihood class, adds integration tests, updates non-bounded parameter implementation by applying bounds from priors and boundary_multiplier argument. Bugfixes to CMAES construction.
• #339 - Updates the calculation of the cyclable lithium capacity in the spme_max_energy example.
• #387 - Adds keys to ParameterSet and updates ECM OCV check.
• #380 - Restore self._boundaries construction for pybop.PSO.
• #372 - Converts np.array to np.asarray for Numpy v2.0 support.
• #165 - Stores the attempted and best parameter values and the best cost for each iteration in the log attribute of the optimiser and updates the associated plots.
• #354 - Fixes the calculation of the gradient in the RootMeanSquaredError cost.
• #347 - Resets options between MSMR tests to cope with a bug in PyBaMM v23.9 which is fixed in PyBaMM v24.1.
• #337 - Restores benchmarks, relaxes CI schedule for benchmarks and scheduled tests.
• #231 - Allows passing of keyword arguments to PyBaMM models and disables build on initialisation.
• #321 - Improves integration/test_spm_parameterisation.py stability, adds flakly pytest plugin, and test_thevenin_parameterisation.py integration test.
• #330 - Fixes implementation of default plotting options.
• #317 - Installs seed packages into nox sessions, ensuring that scheduled tests can pass.
• #308 - Enables testing on both macOS Intel and macOS ARM (Silicon) runners and fixes the scheduled tests.
• #299 - Bugfix multiprocessing support for Linux, MacOS, Windows (WSL) and improves coverage.
• #270 - Updates PR template.
• #91 - Adds a check on the number of parameters for CMAES and makes XNES the default optimiser.

Breaking Changes

• #322 - Add Parameters class to store and access multiple parameters in one object (API change).
• #285 - Drop support for Python 3.8.
• #251 - Drop support for PyBaMM v23.5
• #236 - Restructures the optimiser classes (API change).

v24.3.1 - 2024-06-17

Features

Bug Fixes

• #369 - Upper pins Numpy < 2.0 due to breaking Pints' functionality.

v24.3 - 2024-03-25

Features

• #245 - Updates ruff config for import linting.
• #198 - Adds default subplot trace options, removes [] in axis plots as per SI standard, add varying signal length to quick_plot, restores design optimisation execption.
• #224 - Updated prediction objects to dictionaries, cost class calculations, added additional_variables argument to problem class, updated scipy.minimize defualt method to Nelder-Mead, added gradient cost landscape plots with optional argument.
• #179 - Adds asv configuration for benchmarking and initial benchmark suite.
• #218 - Adds likelihood base class, GaussianLogLikelihoodKnownSigma, GaussianLogLikelihood, and ProbabilityBased cost function. As well as addition of a maximum likelihood estimation (MLE) example.
• #185 - Adds a pull request template, additional nox sessions quick for standard tests + docs, pre-commit for pre-commit, test to run all standard tests, doctest for docs.
• #215 - Adds release_workflow.md and updates release_action.yaml
• #204 - Splits integration, unit, examples, plots tests, update workflows. Adds pytest --examples, --integration, --plots args. Adds tests for coverage after removal of examples. Adds examples and integrations nox sessions. Adds pybop.RMSE._evaluateS1() method
• #206 - Adds Python 3.12 support with corresponding github actions changes.
• #18 - Adds geometric parameter fitting capability, via model.rebuild() with model.rebuild_parameters.
• #203 - Adds support for modern Python packaging via a pyproject.toml file and configures the pytest test runner and ruff linter to use their configurations stored as declarative metadata.
• #123 - Configures scheduled tests to run against the last three PyPI releases of PyBaMM via dynamic GitHub Actions matrix generation.
• #187 - Adds M1 Github runner to test_on_push workflow, updt. self-hosted supported python versions in scheduled tests.
• #118 - Adds example jupyter notebooks.
• #151 - Adds a standalone version of the Problem class.
• #12 - Adds initial implementation of an Observer class and an unscented Kalman filter.
• #190 - Adds a second example design cost, namely the VolumetricEnergyDensity.

Bug Fixes

• #259 - Fix gradient calculation from model.simulateS1 to remove cross-polution and refactor cost._evaluateS1 for fitting costs.
• #233 - Enforces model rebuild on initialisation of a Problem to allow a change of experiment, fixes if statement triggering current function update, updates predictions to simulation to keep distinction between predict and simulate and adds test_changes.
• #123 - Reinstates check for availability of parameter sets via PyBaMM upon retrieval by pybop.ParameterSet.pybamm().
• #196 - Fixes failing observer cost tests.
• #63 - Removes NLOpt Optimiser from future releases. This is to support deployment to the Apple M-Series platform.
• #164 - Fixes convergence issues with gradient-based optimisers, changes default model.check_params() to allow infeasible solutions during optimisation iterations. Adds a feasibility check on the optimal parameters.
• #211 - Allows a subset of parameter bounds or bounds=None to be passed, returning warnings where needed.

v23.12 - 2023-12-19

Features

• #141 - Adds documentation with Sphinx and PyData Sphinx Theme. Updates docstrings across package, relocates costs and dataset to top-level of package. Adds noxfile session and deployment workflow for docs.
• #131 - Adds SciPyDifferentialEvolution optimiser, adds functionality for user-selectable maximum iteration limit to SciPyMinimize, NLoptOptimize, and BaseOptimiser classes.
• #107 - Adds Equivalent Circuit Model (ECM) with examples, Import/Export parameter methods ParameterSet.import_parameter and ParameterSet.export_parameters, updates default FittingProblem.signal definition to "Voltage [V]", and testing infrastructure
• #127 - Adds Windows and macOS runners to the test_on_push action
• #114 - Adds standard plotting class pybop.StandardPlot() via plotly backend
• #114 - Adds quick_plot(), plot_convergence(), and plot_cost2d() methods
• #114 - Adds a SciPy minimize example and logging for non-Pints optimisers
• #116 - Adds PSO, SNES, XNES, ADAM, and IPropMin optimisers to PintsOptimisers() class
• #38 - Restructures the Problem classes ahead of adding a design optimisation example
• #38 - Updates tests and adds a design optimisation example script spme_max_energy
• #120 - Updates the parameterisation test settings including the number of iterations
• #145 - Reformats Dataset to contain a dictionary and signal into a list of strings

Bug Fixes

• #182 - Allow square-brackets indexing of Dataset

v23.11

• Initial release
• Adds Pints, NLOpt, and SciPy optimisers
• Adds SumofSquareError and RootMeanSquareError cost functions
• Adds Parameter and Dataset classes