Birch is run from the command line. Complete installation instructions are available on this web page.
We have used the following versions (referring to particular commits) in our experiments:
Birch: 9fd55ad Birch.Standard: 0c4c453
Once Birch is installed on a computer, our phylogenetics package can be built and installed in the standard manner for a Birch package:
birch build
birch install
To improve the performance it is possible to specify different building options, e.g. --disable-debug to disable debugging. More information on these options can be found in the documentation for the build and the install commands. (Remember to use the same options when you are building the standard library.)
Files in the config directory allow the configuration of various options such as the model, the number of samples to draw, and the number of particles to use. Files in the input directory provide the data sets in an appropriate format for use, including the location of the tree (in the PhyJSON format) as well as the parameters for the prior distributions. Source files for all models can be found in the bi directory.
Inference for the individual models can be run with (e.g. for the CRBD model and the Anatinae tree):
birch sample --config config/crbd.json --input input/anatinae.json --output output/anatinae_crbd.json
Output appears in the output directory, as specified in the command.
Our package also includes the sampler program which stores samples for all particles, rather than just a selected one as the sample program does. To run the included program, change sample to sampler in the above-mentioned command.