https://github.com/sbi-dev/sbi
Simulation-based inference toolkit
https://github.com/sbi-dev/sbi
bayesian-inference likelihood-free-inference machine-learning parameter-estimation pytorch simulation-based-inference
Last synced: 3 days ago
JSON representation
Simulation-based inference toolkit
- Host: GitHub
- URL: https://github.com/sbi-dev/sbi
- Owner: sbi-dev
- License: apache-2.0
- Created: 2020-02-11T13:31:28.000Z (about 5 years ago)
- Default Branch: main
- Last Pushed: 2024-10-19T10:58:04.000Z (7 months ago)
- Last Synced: 2024-10-20T06:36:42.612Z (7 months ago)
- Topics: bayesian-inference, likelihood-free-inference, machine-learning, parameter-estimation, pytorch, simulation-based-inference
- Language: Python
- Homepage: https://sbi-dev.github.io/sbi/
- Size: 95 MB
- Stars: 581
- Watchers: 21
- Forks: 150
- Open Issues: 38
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG.md
- Contributing: CONTRIBUTING.md
- License: LICENSE.txt
- Code of conduct: CODE_OF_CONDUCT.md
- Citation: CITATION.cff
- Codemeta: codemeta.json
Awesome Lists containing this project
- awesome-sciml - mackelab/sbi: Simulation-based inference in PyTorch
README
[](https://badge.fury.io/py/sbi)
[](https://github.com/conda-forge/sbi-feedstock)
[](https://github.com/sbi-dev/sbi/blob/master/CONTRIBUTING.md)
[](https://github.com/sbi-dev/sbi/actions)
[](https://codecov.io/gh/sbi-dev/sbi)
[](https://github.com/sbi-dev/sbi/blob/master/LICENSE.txt)
[](https://doi.org/10.21105/joss.07754)
[](https://numfocus.org/sponsored-projects/affiliated-projects)
## `sbi`: Simulation-Based Inference
[Getting Started](https://sbi-dev.github.io/sbi/latest/tutorials/00_getting_started/) |
[Documentation](https://sbi-dev.github.io/sbi/) | [Discord Server](https://discord.gg/eEeVPSvWKy)
`sbi` is a Python package for simulation-based inference, designed to meet the needs of
both researchers and practitioners. Whether you need fine-grained control or an
easy-to-use interface, `sbi` has you covered.
With `sbi`, you can perform parameter inference using Bayesian inference: Given a
simulator that models a real-world process, SBI estimates the full posterior
distribution over the simulator’s parameters based on observed data. This distribution
indicates the most likely parameter values while additionally quantifying uncertainty
and revealing potential interactions between parameters.
### Key Features of `sbi`
`sbi` offers a blend of flexibility and ease of use:
- **Low-Level Interfaces**: For those who require maximum control over the inference
process, `sbi` provides low-level interfaces that allow you to fine-tune many aspects
of your workflow.
- **High-Level Interfaces**: If you prefer simplicity and efficiency, `sbi` also offers
high-level interfaces that enable quick and easy implementation of complex inference
tasks.
In addition, `sbi` supports a wide range of state-of-the-art inference algorithms (see
below for a list of implemented methods):
- **Amortized Methods**: These methods enable the reuse of posterior estimators across
multiple observations without the need to retrain.
- **Sequential Methods**: These methods focus on individual observations, optimizing the
number of simulations required.
Beyond inference, `sbi` also provides:
- **Validation Tools**: Built-in methods to validate and verify the accuracy of your
inferred posteriors.
- **Plotting and Analysis Tools**: Comprehensive functions for visualizing and analyzing
results, helping you interpret the posterior distributions with ease.
Getting started with `sbi` is straightforward, requiring only a few lines of code:
```python
from sbi.inference import NPE
# Given: parameters theta and corresponding simulations x
inference = NPE(prior=prior)
inference.append_simulations(theta, x).train()
posterior = inference.build_posterior()
```
### Installation
`sbi` requires Python 3.10 or higher. While a GPU isn't necessary, it can improve
performance in some cases. We recommend using a virtual environment with
[`conda`](https://docs.conda.io/en/latest/miniconda.html) for an easy setup.
If `conda` is installed on the system, an environment for installing `sbi` can be created as follows:
```bash
conda create -n sbi_env python=3.10 && conda activate sbi_env
```
### From PyPI
To install `sbi` from PyPI run
```bash
python -m pip install sbi
```
### From conda-forge
To install and add `sbi` to a project with [`pixi`](https://pixi.sh/), from the project directory run
```bash
pixi add sbi
```
and to install into a particular conda environment with [`conda`](https://docs.conda.io/projects/conda/), in the activated environment run
```bash
conda install --channel conda-forge sbi
```
If [`uv`](http://docs.astral.sh/uv/) is installed on the system, an environment for installing `sbi` can be created as follows:
```bash
uv venv -p 3.10
```
Then activate the virtual enviroment by running:
- For `macOS` or `Linux` users
```bash
source .venv/bin/activate
```
- For `Windows` users
```bash
.venv\Scripts\activate
```
To install `sbi` run
```bash
uv add sbi
```
### Testing the installation
Open a Python prompt and run
```python
from sbi.examples.minimal import simple
posterior = simple()
print(posterior)
```
## Tutorials
If you're new to `sbi`, we recommend starting with our [Getting
Started](https://sbi-dev.github.io/sbi/latest/tutorials/00_getting_started/) tutorial.
You can also access and run these tutorials directly in your browser by opening
[Codespace](https://docs.github.com/en/codespaces/overview). To do so, click the green
“Code” button on the GitHub repository and select “Open with Codespaces.” This provides
a fully functional environment where you can explore `sbi` through Jupyter notebooks.
## Inference Algorithms
The following inference algorithms are currently available. You can find instructions on
how to run each of these methods
[here](https://sbi-dev.github.io/sbi/latest/tutorials/16_implemented_methods/).
### Neural Posterior Estimation: amortized (NPE) and sequential (SNPE)
- [`(S)NPE_A`](https://sbi-dev.github.io/sbi/latest/reference/#sbi.inference.trainers.npe.npe_a.NPE_A)
(including amortized single-round `NPE`) from Papamakarios G and Murray I [_Fast
ε-free Inference of Simulation Models with Bayesian Conditional Density
Estimation_](https://proceedings.neurips.cc/paper/2016/hash/6aca97005c68f1206823815f66102863-Abstract.html)
(NeurIPS 2016).
- [`(S)NPE_B`](https://sbi-dev.github.io/sbi/latest/reference/#sbi.inference.trainers.npe.npe_b.NPE_B)
from Lueckmann JM, Goncalves P, Bassetto G, Öcal K, Nonnenmacher M, and Macke J [_Flexible
statistical inference for mechanistic models of neural dynamics_](https://arxiv.org/abs/1711.01861)
(NeurIPS 2017).
- [`(S)NPE_C`](https://sbi-dev.github.io/sbi/latest/reference/#sbi.inference.trainers.npe.npe_c.NPE_C)
or `APT` from Greenberg D, Nonnenmacher M, and Macke J [_Automatic Posterior
Transformation for likelihood-free inference_](https://arxiv.org/abs/1905.07488) (ICML
2019).
- `TSNPE` from Deistler M, Goncalves P, and Macke J [_Truncated proposals for scalable
and hassle-free simulation-based inference_](https://arxiv.org/abs/2210.04815)
(NeurIPS 2022).
- [`FMPE`](https://sbi-dev.github.io/sbi/latest/reference/#sbi.inference.trainers.fmpe.fmpe.FMPE)
from Wildberger, J., Dax, M., Buchholz, S., Green, S., Macke, J. H., & Schölkopf, B.
[_Flow matching for scalable simulation-based
inference_](https://proceedings.neurips.cc/paper_files/paper/2023/hash/3663ae53ec078860bb0b9c6606e092a0-Abstract-Conference.html).
(NeurIPS 2023).
- [`NPSE`](https://sbi-dev.github.io/sbi/latest/reference/#sbi.inference.trainers.npse.npse.NPSE) from
Geffner, T., Papamakarios, G., & Mnih, A. [_Compositional score modeling for
simulation-based inference_](https://proceedings.mlr.press/v202/geffner23a.html).
(ICML 2023)
### Neural Likelihood Estimation: amortized (NLE) and sequential (SNLE)
- [`(S)NLE`](https://sbi-dev.github.io/sbi/latest/reference/#sbi.inference.trainers.nle.nle_a.NLE_A)
or just `SNL` from Papamakarios G, Sterrat DC and Murray I [_Sequential Neural
Likelihood_](https://arxiv.org/abs/1805.07226) (AISTATS 2019).
### Neural Ratio Estimation: amortized (NRE) and sequential (SNRE)
- [`(S)NRE_A`](https://sbi-dev.github.io/sbi/latest/reference/#sbi.inference.trainers.nre.nre_a.NRE_A)
or `AALR` from Hermans J, Begy V, and Louppe G. [_Likelihood-free Inference with
Amortized Approximate Likelihood Ratios_](https://arxiv.org/abs/1903.04057) (ICML
2020).
- [`(S)NRE_B`](https://sbi-dev.github.io/sbi/latest/reference/#sbi.inference.trainers.nre.nre_b.NRE_B)
or `SRE` from Durkan C, Murray I, and Papamakarios G. [_On Contrastive Learning for
Likelihood-free Inference_](https://arxiv.org/abs/2002.03712) (ICML 2020).
- [`(S)NRE_C`](https://sbi-dev.github.io/sbi/latest/reference/#sbi.inference.trainers.nre.nre_c.NRE_C)
or `NRE-C` from Miller BK, Weniger C, Forré P. [_Contrastive Neural Ratio
Estimation_](https://arxiv.org/abs/2210.06170) (NeurIPS 2022).
- [`BNRE`](https://sbi-dev.github.io/sbi/latest/reference/#sbi.inference.trainers.nre.bnre.BNRE) from
Delaunoy A, Hermans J, Rozet F, Wehenkel A, and Louppe G. [_Towards Reliable
Simulation-Based Inference with Balanced Neural Ratio
Estimation_](https://arxiv.org/abs/2208.13624) (NeurIPS 2022).
### Neural Variational Inference, amortized (NVI) and sequential (SNVI)
- [`SNVI`](https://sbi-dev.github.io/sbi/latest/reference/#sbi.inference.posteriors.vi_posterior)
from Glöckler M, Deistler M, Macke J, [_Variational methods for simulation-based
inference_](https://openreview.net/forum?id=kZ0UYdhqkNY) (ICLR 2022).
### Mixed Neural Likelihood Estimation (MNLE)
- [`MNLE`](https://sbi-dev.github.io/sbi/latest/reference/#sbi.inference.trainers.nle.mnle.MNLE) from
Boelts J, Lueckmann JM, Gao R, Macke J, [_Flexible and efficient simulation-based
inference for models of decision-making_](https://elifesciences.org/articles/77220)
(eLife 2022).
## Feedback and Contributions
We welcome any feedback on how `sbi` is working for your inference problems (see
[Discussions](https://github.com/sbi-dev/sbi/discussions)) and are happy to receive bug
reports, pull requests, and other feedback (see
[contribute](https://sbi-dev.github.io/sbi/latest/contribute/)). We wish to maintain a
positive and respectful community; please read our [Code of
Conduct](CODE_OF_CONDUCT.md).
## Acknowledgments
`sbi` is the successor (using PyTorch) of the
[`delfi`](https://github.com/mackelab/delfi) package. It started as a fork of Conor M.
Durkan's `lfi`. `sbi` runs as a community project. See also
[credits](https://github.com/sbi-dev/sbi/blob/master/docs/docs/credits.md).
## Support
`sbi` has been supported by the German Federal Ministry of Education and Research (BMBF)
through project ADIMEM (FKZ 01IS18052 A-D), project SiMaLeSAM (FKZ 01IS21055A) and the
Tübingen AI Center (FKZ 01IS18039A). Since 2024, `sbi` is supported by the appliedAI
Institute for Europe, and by NumFOCUS.
## License
[Apache License Version 2.0 (Apache-2.0)](https://www.apache.org/licenses/LICENSE-2.0)
## Citation
The `sbi` package has grown and improved significantly since its initial release, with
contributions from a large and diverse community. To reflect these developments and the
expanded functionality, we published an [updated JOSS
paper](https://doi.org/10.21105/joss.07754). We encourage you to cite this
newer version as the primary reference:
```latex
@article{BoeltsDeistler_sbi_2025,
doi = {10.21105/joss.07754},
url = {https://doi.org/10.21105/joss.07754},
year = {2025},
publisher = {The Open Journal},
volume = {10},
number = {108},
pages = {7754},
author = {Jan Boelts and Michael Deistler and Manuel Gloeckler and Álvaro Tejero-Cantero and Jan-Matthis Lueckmann and Guy Moss and Peter Steinbach and Thomas Moreau and Fabio Muratore and Julia Linhart and Conor Durkan and Julius Vetter and Benjamin Kurt Miller and Maternus Herold and Abolfazl Ziaeemehr and Matthijs Pals and Theo Gruner and Sebastian Bischoff and Nastya Krouglova and Richard Gao and Janne K. Lappalainen and Bálint Mucsányi and Felix Pei and Auguste Schulz and Zinovia Stefanidi and Pedro Rodrigues and Cornelius Schröder and Faried Abu Zaid and Jonas Beck and Jaivardhan Kapoor and David S. Greenberg and Pedro J. Gonçalves and Jakob H. Macke},
title = {sbi reloaded: a toolkit for simulation-based inference workflows},
journal = {Journal of Open Source Software}
}
```
This updated paper, with its expanded author list, reflects the broader community
contributions and the package's enhanced capabilities in releases
[0.23.0](https://github.com/sbi-dev/sbi/releases/tag/v0.23.3) and later.
If you are using a version of `sbi` prior to 0.23.0, please cite the original sbi
software paper:
```latex
@article{tejero-cantero2020sbi,
doi = {10.21105/joss.02505},
url = {https://doi.org/10.21105/joss.02505},
year = {2020},
publisher = {The Open Journal},
volume = {5},
number = {52},
pages = {2505},
author = {Alvaro Tejero-Cantero and Jan Boelts and Michael Deistler and Jan-Matthis Lueckmann and Conor Durkan and Pedro J. Gonçalves and David S. Greenberg and Jakob H. Macke},
title = {sbi: A toolkit for simulation-based inference},
journal = {Journal of Open Source Software}
}
```
Regardless of which software paper you cite, please also remember to cite the original
research articles describing the specific sbi-algorithm(s) you are using.
Specific releases of `sbi` are also citable via
[Zenodo](https://zenodo.org/records/15034786), where we generate a new software DOI for
each release.