Skip to content

Birch

Birch is a probabilistic programming language featuring automatic marginalization, automatic conditioning, automatic differentiation, and inference algorithms based on Sequential Monte Carlo (SMC). The Birch language transpiles to C++.

See https://birch.sh for a gentle introduction, and https://docs.birch.sh for reference documentation.

lawmurray codecov

License

Birch is open source software. It is licensed under the Apache License, Version 2.0 (the "License"); you may not use it except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0.

Getting started

Binary packages may be available for your system, see the website. If not, or if you have special requirements, you can install Birch from source. This requires:

  • GNU autoconf, automake, libtool, flex, and bison
  • LibYAML
  • Eigen

The following is optional but recommended for significant performance improvements, and will be linked in automatically if found:

All Birch sources are in the same repository. Clone it:

git clone -b stable https://github.com/lawmurray/Birch.git

and change to the Birch directory:

cd Birch

Then proceed as follows. Note special instructions for Mac in step 2. In addition, on Mac, you can typically omit sudo from these commands.

  1. Install MemBirch by running, from within the membirch/ directory:

    ./bootstrap
    ./configure
    make
    sudo make install
    

  2. Install NumBirch by running, from within the numbirch/ directory:

    ./bootstrap
    ./configure
    make
    sudo make install
    

  3. Install Birch by running, from within the birch/ directory:

    ./bootstrap
    ./configure
    make
    sudo make install
    

  4. Install the Birch standard library by running, from within the libraries/Standard/ directory:

    birch build
    sudo birch install
    

This constitutes a basic install. You can inspect the different components for advanced options, such as disabling assertions to improve performance, or building the CUDA backend for NumBirch. You may also like to install other packages in the libraries/ directory. It is not usual to install the packages in the examples/ directory, although you may like to build and run these locally for learning purposes.