Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/uber/orbit

A Python package for Bayesian forecasting with object-oriented design and probabilistic models under the hood.
https://github.com/uber/orbit

arima bayesian bayesian-methods bayesian-statistics changepoint exponential-smoothing forecast forecasting machine-learning orbit probabilistic probabilistic-programming pyro pystan python pytorch regression regression-models stan time-series

Last synced: 4 days ago
JSON representation

A Python package for Bayesian forecasting with object-oriented design and probabilistic models under the hood.

Host: GitHub
URL: https://github.com/uber/orbit
Owner: uber
License: other
Created: 2020-01-07T18:20:37.000Z (about 5 years ago)
Default Branch: dev
Last Pushed: 2024-07-10T23:00:12.000Z (6 months ago)
Last Synced: 2024-10-29T15:18:45.198Z (2 months ago)
Topics: arima, bayesian, bayesian-methods, bayesian-statistics, changepoint, exponential-smoothing, forecast, forecasting, machine-learning, orbit, probabilistic, probabilistic-programming, pyro, pystan, python, pytorch, regression, regression-models, stan, time-series
Language: Python
Homepage: https://orbit-ml.readthedocs.io/en/stable/
Size: 160 MB
Stars: 1,872
Watchers: 35
Forks: 134
Open Issues: 50
Metadata Files:
- Readme: README.md
- Contributing: CONTRIBUTING.md
- License: LICENSE
- Code of conduct: CODE_OF_CONDUCT.md

Awesome Lists containing this project

awesome_time_series_in_python - Orbit
awesome-time-series - orbit - oriented design and probabilistic models under the hood from Uber. (📦 Packages / Python)
awesome-meteo - Orbit
awesome-rainmana - uber/orbit - A Python package for Bayesian forecasting with object-oriented design and probabilistic models under the hood. (Python)
awesome-marketing-machine-learning - orbit
awesome-python-machine-learning-resources - GitHub - 12% open · ⏱️ 17.08.2022): (概率统计)
awesome-list - Orbit - A Python package for Bayesian forecasting with object-oriented design and probabilistic models under the hood. (Linear Algebra / Statistics Toolkit / Statistical Toolkit)

README

        


   

  Join Slack

    |  

  Documentation

    |  

  Blog - Intro

    |  

  Blog - v1.1



![Orbit banner](https://raw.githubusercontent.com/uber/orbit/dev/docs/img/orbit-banner.png)

-------------------------------------------------------------------------------------------------------------------------------------

![GitHub release (latest SemVer)](https://img.shields.io/github/v/release/uber/orbit)

[![PyPI](https://img.shields.io/pypi/v/orbit-ml)][#pypi-package]

[![Build and Test](https://github.com/uber/orbit/actions/workflows/test.yaml/badge.svg?branch=master)](https://github.com/uber/orbit/actions/workflows/test.yaml)

[![Documentation Status](https://readthedocs.org/projects/orbit-ml/badge/?version=stable)](https://orbit-ml.readthedocs.io/en/latest/?badge=latest)

[![PyPI - Python Version](https://img.shields.io/pypi/pyversions/orbit-ml)][#pypi-package]

[![Downloads](https://pepy.tech/badge/orbit-ml)](https://pepy.tech/project/orbit-ml)

[![Conda Recipe](https://img.shields.io/static/v1?logo=conda-forge&style=flat&color=green&label=recipe&message=orbit-ml)][#conda-forge-feedstock]

[![Conda - Platform](https://img.shields.io/conda/pn/conda-forge/orbit-ml?logo=anaconda&style=flat)][#conda-forge-package]

[![Conda (channel only)](https://img.shields.io/conda/vn/conda-forge/orbit-ml?logo=anaconda&style=flat&color=orange)][#conda-forge-package]

[![PyPI - License](https://img.shields.io/pypi/l/orbit-ml?logo=pypi&style=flat&color=green)][#github-license]

[#github-license]: https://github.com/uber/orbit/blob/master/LICENSE

[#pypi-package]: https://pypi.org/project/orbit-ml/

[#conda-forge-package]: https://anaconda.org/conda-forge/orbit-ml

[#conda-forge-feedstock]: https://github.com/conda-forge/orbit-ml-feedstock

# User Notice

The default page of the repo is on `dev` branch. To install the dev version, please check the section `Installing from Dev Branch`. If you are looking for a **stable version**, please refer to the `master` branch [here](https://github.com/uber/orbit/tree/master).

# Disclaimer

This project

- is stable and being incubated for long-term support. It may contain new experimental code, for which APIs are subject to change.

- requires [cmdstanpy](https://mc-stan.org/cmdstanpy/) as one of the core dependencies for Bayesian sampling.

# Orbit: A Python Package for Bayesian Forecasting

Orbit is a Python package for Bayesian time series forecasting and inference. It provides a

familiar and intuitive initialize-fit-predict interface for time series tasks, while utilizing probabilistic programming languages under the hood.

For details, check out our documentation and tutorials:

- HTML (stable): https://orbit-ml.readthedocs.io/en/stable/

- HTML (latest): https://orbit-ml.readthedocs.io/en/latest/

Currently, it supports concrete implementations for the following models:

-  Exponential Smoothing (ETS)

-  Local Global Trend (LGT)

-  Damped Local Trend (DLT)

-  Kernel Time-based Regression (KTR)

It also supports the following sampling/optimization methods for model estimation/inferences:

-  Markov-Chain Monte Carlo (MCMC) as a full sampling method

-  Maximum a Posteriori (MAP) as a point estimate method

-  Variational Inference (VI) as a hybrid-sampling method on approximate

   distribution

##  Installation

### Installing Stable Release

Install the library either from PyPi or from the source with `pip`. 

Alternatively, you can also install it from Anaconda with `conda`:

**With pip**

1. Installing from PyPI

   ```sh

   $ pip install orbit-ml

   ```

2. Install from source

   ```sh

   $ git clone https://github.com/uber/orbit.git

   $ cd orbit

   $ pip install -r requirements.txt

   $ pip install .

   ```

**With conda**

The library can be installed from the conda-forge channel using conda.

```sh

$ conda install -c conda-forge orbit-ml

```

### Installing from Dev Branch

```sh

$ pip install git+https://github.com/uber/orbit.git@dev

```

## Quick Start with Damped-Local-Trend (DLT) Model

### FULL Bayesian Prediction

```python

from orbit.utils.dataset import load_iclaims

from orbit.models import DLT

from orbit.diagnostics.plot import plot_predicted_data

# log-transformed data

df = load_iclaims()

# train-test split

test_size = 52

train_df = df[:-test_size]

test_df = df[-test_size:]

dlt = DLT(

  response_col='claims', date_col='week',

  regressor_col=['trend.unemploy', 'trend.filling', 'trend.job'],

  seasonality=52,

)

dlt.fit(df=train_df)

# outcomes data frame

predicted_df = dlt.predict(df=test_df)

plot_predicted_data(

  training_actual_df=train_df, predicted_df=predicted_df,

  date_col=dlt.date_col, actual_col=dlt.response_col,

  test_actual_df=test_df

)

```

![full-pred](docs/img/dlt-mcmc-pred.png)

## Demo

Nowcasting with Regression in DLT:

[![Open All Collab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/edwinnglabs/ts-playground/blob/master/Orbit_Tutorial.ipynb)

Backtest on M3 Data:

[![Open All Collab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/edwinnglabs/ts-playground/blob/master/orbit_m3_backtest.ipynb)

More examples can be found under [tutorials](./docs/tutorials)

and [examples](./examples).

# Contributing

We welcome community contributors to the project. Before you start, please read our

[code of conduct](CODE_OF_CONDUCT.md) and check out

[contributing guidelines](CONTRIBUTING.md) first.

# Versioning

We document versions and changes in our [changelog](./docs/changelog.rst).

# References

## Presentations

Check out the ongoing [deck](https://docs.google.com/presentation/d/1WfTtXAW3rud4TX9HtB3NkE6buDE8tWk6BKZ2hRNXjCI/edit?usp=sharing) for scope and roadmap of the project. An older deck used in the [meet-up](https://www.meetup.com/UberEvents/events/279446143/) during July 2021 can also be found [here](https://docs.google.com/presentation/d/1R0Ol8xahIE6XlrAjAi0ewu4nRxo-wQn8w6U7z-uiOzI/edit?usp=sharing).

## Citation

To cite Orbit in publications, refer to the following whitepaper:

[Orbit: Probabilistic Forecast with Exponential Smoothing](https://arxiv.org/abs/2004.08492)

Bibtex:

```

@misc{

    ng2020orbit,

    title={Orbit: Probabilistic Forecast with Exponential Smoothing},

    author={Edwin Ng,

        Zhishi Wang,

        Huigang Chen,

        Steve Yang,

        Slawek Smyl},

    year={2020}, eprint={2004.08492}, archivePrefix={arXiv}, primaryClass={stat.CO}

}

```

##  Papers

- Bingham, E., Chen, J. P., Jankowiak, M., Obermeyer, F., Pradhan, N., Karaletsos, T., Singh, R., Szerlip,

  P., Horsfall, P., and Goodman, N. D. Pyro: Deep universal probabilistic programming. The Journal of Machine Learning

  Research, 20(1):973–978, 2019.

- Hoffman, M.D. and Gelman, A. The No-U-Turn sampler: adaptively setting path lengths in Hamiltonian Monte Carlo.

  J. Mach. Learn. Res., 15(1), pp.1593-1623, 2014.

- Hyndman, R., Koehler, A. B., Ord, J. K., and Snyder, R. D. Forecasting with exponential smoothing:

  the state space approach. Springer Science & Business Media, 2008.

- Smyl, S. Zhang, Q. Fitting and Extending Exponential Smoothing Models with Stan.

  International Symposium on Forecasting, 2015.

## Related projects

- [Pyro](https://github.com/pyro-ppl/pyro)

- [Stan](https://github.com/stan-dev/stan)

- [Rlgt](https://cran.r-project.org/web/packages/Rlgt/index.html)

- [forecast](https://github.com/robjhyndman/forecast)

- [prophet](https://facebook.github.io/prophet/)