Skip to content

TomHochsprung/UAI2023_Pairwise_CI_Testing

Repository files navigation

UAI2023_Pairwise_CI_Testing

This repository contains the R-scripts that were used to generate the plots in the paper that will be submitte to UAI2023 (exact reference to follow). This repository does not constitute a package, it just consists of some scripts.

Requirements

The R version should match the mentioned packages below, that means that the R version should be >= 3.5. (We used R version 3.6.3 (2020-02-29) -- "Holding the Windsock"). Three packages need to be installed in the respective R environment. The user can run the script install_packages.R which should install all the relevant packages. All relevant packages have a general public license. The packages are:

  • mvtnorm (purpose: to generate multivariate normal data, license: GPL-2, version: 1.1-3)
  • energy (purpose: needed for the partial distance correlation, license: GPL-2 | GPL-3, version: 1.7-11)
  • GeneralisedCovarianceMeasure: is only needed when using the Generalized Covariance Measure (purpose: to have an implementation of the gcm test statistic, license: GPL-2, version: 0.2.0)

There is a secondary install packages script, called install_packages_plotting_only.R These packages are only required if the user wants to achieve the same plots as we have, for the execution of the other scripts, this installation is not necessary.

  • ggplot2 (purpose: to generate plots, license: MIT open source, version: 3.4.1)
  • dplyr (purpose: basic data cleaning functionality, license: MIT open source, version: 1.1.0)
  • tidyr (purpose: basic data cleaning functionality, license: MIT open source, version: 1.3.0)
  • ggpubr (purpose: grouping different plots together in one plot, license: GPL-2 | GPL-3, version: 0.6.0)
  • latex2exp (purpose: to use latex expressions in plots, license: MIT open source, version: 0.9.6)
  • lattice (purpose: to make 3d plots, license: GPL-2 | GPL-3, version: 0.20.40)

Contact info

tom dot hochsprung at dlr dot de

Scripts

The repository contains several scripts:

main_code.R: This is the main script of the repository, here the user can change the parameters and the model generates the data.

model_sigma_1.R: This is the model corresponding to Sigma^(1) in the paper.

model_sigma_2.R: This is the model corresponding to Sigma^(2) in the paper.

model_sigma_gcm.R: This is the model corresponding to the simulations regarding the Generalized Covariance Measure in the Appendix.

execute_standard_test.R: This function executes the standard pairwise testing algorithm, see Section 3.1 in the paper.

execute_novel_test.R: This function executes the novel pairwise testing algorithm, see Sections 3.2 and 3.3 in the paper.

test_ci_by_par_cor.R: Tests (conditional) independence using the partial correlation and the z-transform

plotting.R Script to generate plots as we have used.

install_packages.R: Install main important packages.

install_packages_plotting_only.R: Installs packages that are only relevant for plotting as we have done it

theoretical_lower_bounds_2.R: Code to generate the 3d-plot on the left-hand side of Figure 2

theoretical_lower_bounds_3.R: Code to generate the 3d-plot on the right-hand side of Figure 3

License

You can redistribute and/or modify the code under the terms of the GNU General Public License as published by the Free Software Foundation; either version 3 of the License, or (at your option) any later version.

Project status

The code was meant to provide a way of validating the generated plots in the original paper. As of now, no further developments are planned.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages