Statistical hypothesis testing methods for non-parametric functional dependencies using asymptotic chi-square or exact statistics. These tests were motivated to reveal evidence for causality based on functional dependencies. They include asymptotic functional chi-square tests, an exact functional test, a comparative functional chi-square test, and also a comparative chi-square test. The normalized non-constant functional chi-square test was used by Best Performer NMSUSongLab in HPN-DREAM (DREAM8) Breast Cancer Network Inference Challenges. For continuous data, these tests offer an advantage over regression analysis when a parametric functional form cannot be assumed; for categorical data, they provide a novel means to assess directional dependencies not possible with symmetrical Pearson's chi-square or Fisher's exact tests.
| Version: | 2.3.3 |
| Depends: | R (≥ 3.0.0) |
| Imports: | Rcpp, stats |
| LinkingTo: | BH, Rcpp |
| Suggests: | Ckmeans.1d.dp, testthat, knitr, rmarkdown |
| Published: | 2016-09-02 |
| Author: | Yang Zhang, Hua Zhong, and Joe Song |
| Maintainer: | Joe Song <joemsong at cs.nmsu.edu> |
| License: | LGPL (≥ 3) |
| URL: | http://www.cs.nmsu.edu/~joemsong/publications |
| NeedsCompilation: | yes |
| Citation: | FunChisq citation info |
| Materials: | NEWS |
| CRAN checks: | FunChisq results |
| Reference manual: | FunChisq.pdf |
| Vignettes: |
Using the exact functional test Which statistic to use for functional dependency? |
| Package source: | FunChisq_2.3.3.tar.gz |
| Windows binaries: | r-devel: FunChisq_2.3.3.zip, r-release: FunChisq_2.3.3.zip, r-oldrel: FunChisq_2.3.3.zip |
| OS X Mavericks binaries: | r-release: FunChisq_2.3.3.tgz, r-oldrel: FunChisq_2.3.3.tgz |
| Old sources: | FunChisq archive |
Please use the canonical form https://CRAN.R-project.org/package=FunChisq to link to this page.