Tue 30 Jun 2020 01:14:13 PM MDT

1. Please download conda if you don't have it yet. This tool will allow you to
easily download all of the necessary dependencies. You can get it from this
page:

https://docs.conda.io/projects/conda/en/latest/user-guide/install/download.html

2. Enter the directory in which this file resides

3. Use conda to create an environment with all of the necessary dependencies
(listed in boat_env.yml):

   $ conda env create -f boat_env.yml

4. Activate the environment:

   $ conda activate boat_env

5. Run the code with a small (<6) number of pulses!

   $ python run_analysis.py <# of pulses>

   (e.g. $ python run_analysis.py 2)

Note: this program will use all of your cores and could make your system
sluggish. To avoid this, prepend the above with the "nice" command, which
will run the code with lower priority:

   $ nice python run_analysis.py <# of pulses>

6. A summary of the analysis will be generated to a file called

   all_analyses_<# of pulses>pulses_<signature>.<signature>_output.dat

   (e.g. all_analyses_0pulses_1593543779.9746473_output.dat)

The signature changes from run to run, and is there for avoiding overwriting
files.

7. For each trait for which the pulse model is preferred (DeltaAICc < -2), the
code generates figures (with the file extension .png) similar to the ones
described in the manuscript SI. Each figure's filename contains the signature
of the summary file.

8. All of the data used in the analysis is stored as a Python pickle file with file
extension .p. You can safely remove it, but please send it over if you find a bug!

9. Please email us if you have any questions.