3. Installation

You should have Python 2.7 installed.

First, if you have Git installed, clone the Git repository:

$ git clone https://github.com/joaks1/PyMsBayes.git

If you do not have Git, you can download a snapshot of the repository as a tar or zip archive by clicking the respective folder icons at the top of this page. Either way, move into the downloaded directory:

$ cd PyMsBayes

Next, let’s make sure the bundled dpp-msbayes tools work on your machine:

$ python setup.py test

This will run a small number of tests to ensure the tools bundled with PyMsBayes work on your computer. If these tests pass, you are good to go and can install PyMsBayes:

$ sudo python setup.py install

If the install was successful, you should be able to call up the help menu of the main program of PyMsBayes:

$ dmc.py -h

You should see output that begins like:

usage: dmc.py [-h] -o OBSERVED_CONFIGS [OBSERVED_CONFIGS ...] -p PRIOR_CONFIGS
              [PRIOR_CONFIGS ...] [-r REPS] [-n NUM_PRIOR_SAMPLES]
              [--prior-batch-size PRIOR_BATCH_SIZE] [--generate-samples-only]
              [--num-posterior-samples NUM_POSTERIOR_SAMPLES]
              [--num-standardizing-samples NUM_STANDARDIZING_SAMPLES]
              [--np NP] [--output-dir OUTPUT_DIR] [--temp-dir TEMP_DIR]
              [--staging-dir STAGING_DIR]
              [-s [STAT_PREFIXES [STAT_PREFIXES ...]]] [-b BANDWIDTH]
              [-q NUM_POSTERIOR_QUANTILES]
              [--reporting-frequency REPORTING_FREQUENCY]
              [--sort-index {0,1,2,3,4,5,6,7,8,9,10,11}]
              [--no-global-estimate] [--compress] [--keep-temps] [--seed SEED]
              [--output-prefix OUTPUT_PREFIX] [--data-key-path DATA_KEY_PATH]
              [--start-from-simulation-index START_FROM_SIMULATION_INDEX]
              [--start-from-observed-index START_FROM_OBSERVED_INDEX]
              [--dry-run] [--version] [--quiet] [--debug]

dmc.py Version 0.2.4

optional arguments:
  -h, --help            show this help message and exit
  -o OBSERVED_CONFIGS [OBSERVED_CONFIGS ...], --observed-configs OBSERVED_CONFIGS [OBSERVED_CONFIGS ...]
                        One or more msBayes config files to be used to either
                        calculate or simulate observed summary statistics. If
                        used in combination with `-r` each config will be used
                        to simulate pseudo-observed data. If analyzing real
                        data, do not use the `-r` option, and the fasta files
                        specified within the config must exist and contain the
                        sequence data.

3.1. Troubleshooting

If any of the tests fail and/or you get the following error message when you try to install PyMsBayes:

**********************************************************************
****************************** WARNING *******************************
The bundled `dpp-msbayes` tools are not being installed, because some
of them are not executable on this system. The `PyMsBayes` package and
scripts will still be installed, however, you will need to build and
install `dpp-msbayes` (https://github.com/joaks1/dpp-msbayes) yourself
in order to use them.  Sorry for the inconvenience.
**********************************************************************

You will need to build and install dpp-msbayes from https://github.com/joaks1/dpp-msbayes.git, and install ABCtoolbox from http://www.cmpg.iee.unibe.ch/content/softwares__services/computer_programs/abctoolbox/index_eng.html. After you install dpp-msbayes and ABCtoolbox to your PATH, if you already ran the sudo python setup.py install command above, you are good to go. If you haven’t, cd back to the PyMsBayes directory and run this command (Note: you will still get the above warning message, but you can ignore it because the dpp-msbayes tools are already on your system).