Dunn's test computes stochastic dominance & reports pairwise comparisons. This is done following a Kruskal-Wallis test (Kruskal and Wallis, 1952). It employs Dunn's z-test-statistic approximations for rank statistics, conducting k(k-1)/2 comparisons. The null hypothesis assumes that the probability of a randomly selected value from the first group being larger than one from the second group is one half, similar to the Wilcoxon-Mann-Whitney test. Dunn's test serves as a test for median difference and takes into account tied ranks.

The Structural Topic Model (STM) allows researchers to estimate topic models with document-level covariates. The package also includes tools for model selection, visualization, and estimation of topic-covariate regressions.

This package provides implementations of a family of Lasso variants including Dantzig Selector, LAD Lasso, SQRT Lasso, Lq Lasso for estimating high dimensional sparse linear models.

This package provides fast and memory-friendly tools for text vectorization, topic modeling (LDA, LSA), word embeddings (GloVe), similarities. It provides a source-agnostic streaming API, which allows researchers to perform analysis of collections of documents which are larger than available RAM. All core functions are parallelized to benefit from multicore machines.

Pdist computes the euclidean distance between rows of a matrix X and rows of another matrix Y. Previously, this could be done by binding the two matrices together and calling dist, but this creates unnecessary computation by computing the distances between a row of X and another row of X, and likewise for Y. Pdist strictly computes distances across the two matrices, not within the same matrix, making computations significantly faster for certain use cases.

This package can be used to compute local false discovery rates.

The main purpose of this package is to provide the algorithmic complexity for short strings, an approximation of the Kolmogorov Complexity of a short string using the coding theorem method. While the database containing the complexity is provided in the data only package acss.data, this package provides functions accessing the data such as prob_random returning the posterior probability that a given string was produced by a random process. In addition, two traditional (but problematic) measures of complexity are also provided: entropy and change complexity.

This package computes various confidence intervals (CI) for the Kaplan-Meier estimator, namely: Petos CI, Rothman CI, CIs based on Greenwoods variance, Thomas and Grunkemeier CI and the simultaneous confidence bands by Nair and Hall and Wellner.

This package implements reinforcement learning environments and algorithms as described in Sutton & Barto (1998). The Q-Learning algorithm can be used with function approximation, eligibility traces (Singh & Sutton, 1996) and experience replay (Mnih et al., 2013).

This package provides an optimization method based on sequential quadratic programming for maximum likelihood estimation of the mixture proportions in a finite mixture model where the component densities are known. The algorithm is expected to obtain solutions that are at least as accurate as the state-of-the-art MOSEK interior-point solver, and they are expected to arrive at solutions more quickly when the number of samples is large and the number of mixture components is not too large.

This package provides a system for reporting messages, which offers certain useful features over the standard R system, such as the incorporation of output consolidation, message filtering, assertions, expression substitution, automatic generation of stack traces for debugging, and conditional reporting based on the current "output level".

High dimensional interaction search by brute force requires a quadratic computational cost in the number of variables. The xyz algorithm provably finds strong interactions in almost linear time. For details of the algorithm see: G. Thanei, N. Meinshausen and R. Shah (2016). The xyz algorithm for fast interaction search in high-dimensional data.

This package lets you determine the significance of pre-defined sets of genes with respect to an outcome variable, such as a group indicator, a quantitative variable or a survival time.

This package provides functions to fit kernel density functions to animal activity time data; plot activity distributions; quantify overall levels of activity; statistically compare activity metrics through bootstrapping; and evaluate variation in linear variables with time (or other circular variables).

This package lets you generate random colors, possibly with a given hue or a given luminosity.

This package provides alternative implementations of some base R functions, including sort, order, and match. The functions are simplified but can be faster or have other advantages.

This package provides a compilation of extra ggplot2 themes, scales and utilities, including a spell check function for plot label fields and an overall emphasis on typography.

Fit Bayesian generalized (non-)linear multivariate multilevel models using Stan for full Bayesian inference. A wide range of distributions and link functions are supported, allowing users to fit -- among others -- linear, robust linear, count data, survival, response times, ordinal, zero-inflated, hurdle, and even self-defined mixture models all in a multilevel context. Further modeling options include non-linear and smooth terms, auto-correlation structures, censored data, meta-analytic standard errors, and quite a few more. In addition, all parameters of the response distribution can be predicted in order to perform distributional regression. Prior specifications are flexible and explicitly encourage users to apply prior distributions that actually reflect their beliefs. Model fit can easily be assessed and compared with posterior predictive checks and leave-one-out cross-validation.

Sending functions to remote processes can be wasteful of resources because they carry their environments with them. With this package, it is easy to create functions that are isolated from their environment. These isolated functions, also called crates, print to the console with their total size and can be easily tested locally before being sent to a remote.

This package provides functions for kernel-regression-based association tests including Burden test, SKAT and SKAT-O. These methods aggregate individual SNP score statistics in a SNP set and efficiently compute SNP-set level p-values.

This package lets you interact with Google Sheets through the Sheets API v4. This package can read and write both the metadata and the cell data in a Sheet.

This package provides support for simple features, a standardized way to encode spatial vector data. It binds to GDAL for reading and writing data, to GEOS for geometrical operations, and to PROJ for projection conversions and datum transformations.

This package provides a fast match replacement for cases that require repeated look-ups. It is slightly faster that R's built-in match function on first match against a table, but extremely fast on any subsequent lookup as it keeps the hash table in memory.

This package implements a data structure similar to hashes in Perl and dictionaries in Python but with a purposefully R flavor. For objects of appreciable size, access using hashes outperforms native named lists and vectors.