Wagner

A theoretical model to study macroevolution and phylogeographies.

Named after Moritz Wagner (1813-1887), who made important contributions to our understanding of speciation.

The program is written in C++14. Tested on Linux 64-bit with clang++ 4.0 and g++ 6.3.0.

The program outputs xml files for info and a graphml file for the network of communities. The python folder has scripts to extract information from the raw results.

Reference

Working paper: http://arxiv.org/abs/1203.1790

Dependencies

boost [optional]. Wagner defaults to the much slower std::set/std::map if boost is unavailable.
cmake for building

usage and options

Compile with make (requires a modern ANSI C++14 compiler), e.g.:

$ mkdir build && cd $_
$ cmake ..
$ make

and execute with

$ ./src/wagner_exe

You can change the model with:

-model

The current version supports four models:

0   Neutral model
1   Phylogenetic distance
2   Euclidean distance with traits [default]
3   Fuzzy distance with traits

You can also use the following options [default values]:

-threads    Number of threads to launch [number of available cores].
-seed       Seed for the random number generator [6].
-c          Number of communities (or vertices, or nodes, or patches) [64].
-t          Number of time steps. Needs to be a power of two [512].
-e          Per population extinction rate [0.05].
-m          Max migration rate [0.04].
-n          Number of traits for models 2-3 [10].
-w          Standard deviation of the white noise applied to traits after each time step [0.001].
-a          Aleph for models 1-2 [10.0].
-s          Speciation rate [0.04].
-r          Radius of the random geometric network [0.2].

Options not followed by an argument

-shuffle    After t/2 time steps, shuffle all populations [false].

For example:

$ ./ẃagner -model 1 -a 50 -seed 88

Will run the model 1 with aleph = 50, and seed = 88. The order of the options doesn't matter and all parameters should be printed in the xml output file. If the number of time steps supplied is not a power of two, the problem will find the largest power of two that fits in this number.

license

MIT http://opensource.org/licenses/MIT

author

Philippe Desjardins-Proulx philippe.d.proulx@gmail.com

PhDP / Wagner