Skip to contents

phyopt

This packages implements a two-step, population-based, hybrid optimization algorithm, for solving problems of the form:

\[ \begin{array}{c} \max_{x, \tilde{x}} f(x, \tilde{x}) ~~s.t.~~ g(x, \tilde{x}) \geq 0,~ \tilde{g}(\tilde{x}) \geq 0,\\ x \in X,~ \tilde{x} \in \tilde{X} \end{array} \]

The paper fully describing the motivation, theoretical framework, and method design is available at vignette("theoretical_framework", package = "phyopt").

For now, see the vignette vignette("example", package = "phyopt"), which contains a simple example but a thorough explanation of how the package works. Additionally, see the documentation for the optimize_phy() function, which is the main function of the package

Disclaimer: this package is in the early most stage of life. It hasn’t been thoroughly tested and can present several bugs. I don’t recommend using it for large-scale projects, yet.

Please report any problems as a GitHub issue. Other comments can be posted as a GitHub discussion, or sent in my email below. Thank you!

Author: Ricardo Semião e Castro ().

Installation

You can install the development version of phyopt like so:

# install.packages("devtools")
devtools::install_github("ricardo-semiao/phyopt")

Development Comments

This package is young, and puts much of the responsibility of the optimization on the user. Having said that, I think it is decently solid. Once I implement tests and run more examples, I figure it will be already stable.

Some of the more important features I plan to add in the future are:

  • The testthat package structure is created for the package, but I haven’t had the time to write unir tests yet.
  • Add more complicated examples to vignette("example", package = "phyopt").
  • Expand the suggested metrics on get_metrics, also add suggestions formatters for the logger object.
  • Add suggested operators for optimize_phy(), with common initializers, optimizers, and updaters from the literature.
  • Add an on.exit() call to optimize_pbhy(), such that it safely returns what it had before reaching an eventual error.
  • Create summary() and plot() methods for the optimize_pbhy() result.

Note that this package: