GPTIPS 2F Toolbox for MATLAB

GPTIPS 2F tool­box for MATLAB is a fork of the orig­i­nal GPTIPS2 tool­box devel­oped by Dominic Sear­son. The acronym stands for “Genet­ic Pro­gram­ming Tool­box for the Iden­ti­fi­ca­tion of Phys­i­cal Sys­tems” and “2F” refers to the fact that it is both a fork (of ver­sion 2 of the tool­box) and a func­tion­al update.

That said, the updat­ed tool­box deliv­ers the fol­low­ing functionality:

  1. Pre­set Ran­dom Con­stants (PRCs): a sub­set of Ephemer­al Ran­dom Con­stants (ERCs) the dif­fer­ence being that PRCs are ran­dom­ly cho­sen from a pre­de­fined set.
  2. Auto­mat­i­cal­ly Defined Func­tions (ADFs): basi­cal­ly tem­plates that are seed­ed into the ini­tial pop­u­la­tion and can also arise nat­u­ral­ly dur­ing mutation.
  3. Evo­lu­tion­ary rules: define rules for either out­right killing indi­vid­u­als with cer­tain unde­sir­able traits or sig­nif­i­cant­ly decrease their chances of sur­vival (alpha test­ing feature).

Addi­tion­al­ly, a num­ber of bugs were fixed and a seri­ous com­pat­i­bil­i­ty issue was resolved for Mat­lab R2018a/b+.

The fork is large­ly a work-in-progress (although it is already ful­ly usable) with con­tri­bu­tions from sev­er­al Ph.D. stu­dents grad­u­al­ly being inte­grat­ed into the toolbox.