'o' (optimize) method has been removed as the benchmarking itself proved to be prohibitively slow. Instead, the default behavior now is to infer which method is likely faster based on the sample sizes and dimension of the data.cores argument has been removed. Use threads instead.std::vector was used to tabulate the number of points in each octant, and the vector is of size 2^d. For larger dimensions, tabulation is now done using a std::unordered_map, the size of which scales with the sample sizes instead of dimension.seed argument is no longer ignored when threads is larger than 1.'o' for optimize: the test statistic is computed using both the range tree and brute force methods, and the faster of the two is used for the permutation test.cores argument has been renamed threads. cores is still allowed for backwards compatibility but will be removed in the next release.seed argument has been added to ensure reproducibility in the p-value estimation.verbose argument has been added to print a progress bar during the permutation test.nBootstrap argument is deprecated, and will be removed in the next version of this package. Use nPermute instead of nBootstrap to compute the empirical null distribution for your data.quadCount() calculation reducing runtime.Htest for improved integration with other packages.The fasano.franceschini.test has been released into the wild!
fasano.franceschini.test, see the accompanying paper.Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.