FunChisq: Chi-Square and Exact Tests for Non-Parametric Functional
Dependencies
Statistical hypothesis testing methods for non-parametric functional dependencies based on asymptotic chi-square and exact statistics. These tests were designed 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 the Best Performer NMSUSongLab in HPN-DREAM (DREAM8) Breast Cancer Network Inference Challenges. These tests offer an advantage over regression analysis when a parametric functional form cannot be assumed.
Downloads: