brian2modelfitting

Model fitting toolbox for Brian 2 simulator.

The package brian2modelfitting allows the user to find the best fit of the unknown free parameters for recorded traces and spike trains. It also supports simulation-based inference, where instead of point-estimated parameter values, a full posterior distribution over the parameters is computed.

By default, the toolbox supports a range of global derivative-free optimization methods, that include popular methods for model fitting: differential evolution, particle swarm optimization and covariance matrix adaptation (provided by the Nevergrad, a gradient-free optimization platform) as well as Bayesian optimization for black box functions (provided by scikit-optimize, a sequential model-based optimization library). On the other hand, simulation-based inference is the process of finding parameters of a simulator from observations by taking a Bayesian approach via sequential neural posterior estimation, likelihood estimation or ration estimation (provided by the sbi), where neural densitiy estimator, a deep neural network allowing probabilistic association between the data and underlying parameter space, is trained. After the network is trained, the approximated posterior distribution is available.

Documentation

The full documentation is available at http://brian2modelfitting.readthedocs.org.

Testing status

Installation

Install brian2modelfitting from the Python package index via pip:

pip install brian2modelfitting

The basic install only supports the optimization methods provided by the Nevergrad library. To install additional algorithms, support for simulation-based inference, and other features, you can install the optional dependencies:

| extra | description | install command | |---------|---------------------------------------------------------------------------------------|-------------------------------------------| | sbi | simulation-based inference with sbi | pip install 'brian2modelfitting[sbi]' | | skopt | optimization with sckikit-optimize | pip install 'brian2modelfitting[skopt]' | | algos | additional algorithms for Nevergrad | pip install 'brian2modelfitting[algos]' | | efel | FeatureMetric with eFEL electrophysiology features | pip install 'brian2modelfitting[efel]' | | test | framework to run the test suite | pip install 'brian2modelfitting[test]' | | docs | framework to build the documentation | pip install 'brian2modelfitting[docs]' | | all | all of the above dependencies | pip install 'brian2modelfitting[all]' |

License
-------
The model fitting toolbox is released under the terms of the CeCILL 2.1 license and is available [here](https://github.com/brian-team/brian2modelfitting/blob/master/LICENSE).

Use
---
Please report issues at the GitHub [issue tracker](https://github.com/brian-team/brian2modelfitting/issues) or at the [Brian 2 discussion forum](https://brian.discourse.group).