This package provides randomization tests and graphical diagnostics for assessing randomized assignment and covariate balance for a binary treatment variable. See Branson (2021) <arXiv:1804.08760> for details.

This package provides a framework for unit testing for realistic minimalists, where we distinguish between expected, acceptable, current, fallback, ideal, or regressive behaviour. It can also be used for monitoring third-party software projects for changes.

Foundational package in the R4SUB (R for Regulatory Submission) ecosystem. Defines the core evidence table schema, parsers, indicator abstractions, and scoring primitives needed to quantify clinical submission readiness. Provides a standardized contract for ingesting heterogeneous sources (validation outputs, metadata, traceability) into a single evidence framework.

This package provides a Tidy implementation of grouping sets', rollup and cube - extensions of the group_by clause that allow for computing multiple group_by clauses in a single statement. For more detailed information on these functions, please refer to "Enhanced Aggregation, Cube, Grouping and Rollup" <https://cwiki.apache.org/confluence/display/Hive/Enhanced+Aggregation%2C+Cube%2C+Grouping+and+Rollup>.

Mappable vector library provides convenient way to access large datasets. Use all of your data at once, with few limits. Memory mapped data can be shared between multiple R processes. Access speed depends on storage medium, so solid state drive is recommended, preferably with PCI Express (or M.2 nvme) interface or a fast network file system. The data is memory mapped into R and then accessed using usual R list and array subscription operators. Convenience functions are provided for merging, grouping and indexing large vectors and data.frames. The layout of underlying MVL files is optimized for large datasets. The vectors are stored to guarantee alignment for vector intrinsics after memory map. The package is built on top of libMVL, which can be used as a standalone C library. libMVL has simple C API making it easy to interchange datasets with outside programs. Large MVL datasets are distributed via Academic Torrents <https://academictorrents.com/collection/mvl-datasets>.

This package provides a set of tools to streamline data analysis. Learning both R and introductory statistics at the same time can be challenging, and so we created rigr to facilitate common data analysis tasks and enable learners to focus on statistical concepts. We provide easy-to-use interfaces for descriptive statistics, one- and two-sample inference, and regression analyses. rigr output includes key information while omitting unnecessary details that can be confusing to beginners. Heteroscedasticity-robust ("sandwich") standard errors are returned by default, and multiple partial F-tests and tests for contrasts are easy to specify. A single regression function can fit both linear and generalized linear models, allowing students to more easily make connections between different classes of models.

An implementation of a stochastic heuristic method for performing multidimensional function optimization. The method is inspired in the Cross-Entropy Method. It does not relies on derivatives, neither imposes particularly strong requirements into the function to be optimized. Additionally, it takes profit from multi-core processing to enable optimization of time-consuming functions.

This package provides XML parsing capability through the Rapidxml C++ header-only library.

This package provides estimation and data generation tools for several new regression models, including the gamma, beta, inverse gamma, beta prime, log-normal and log-logistic distributions. These models can be parameterized based on the mean, median, mode, geometric mean and harmonic mean, except for the log-logistic model which is based on alternative parametrizations. For details, see Bourguignon and Gallardo (2025a) <doi:10.1016/j.chemolab.2025.105382> and Bourguignon and Gallardo (2025b) <doi:10.1111/stan.70007>. The package also implements higher-order likelihood inference through Skovgaard-adjusted likelihood ratio statistics and predictive shrinkage estimators reparameterized beta regression models.

Slow Feature Analysis (SFA), ported to R based on matlab implementations of SFA: SFA toolkit 1.0 by Pietro Berkes and SFA toolkit 2.8 by Wolfgang Konen.

R6 class interface for handling relational database connections using DBI package as backend. The class allows handling of connections to e.g. PostgreSQL, MariaDB and SQLite. The purpose is having an intuitive object allowing straightforward handling of SQL databases.

Analytical computation of rolling and expanding Shapley values for time-series data. The rollshap package decomposes the coefficient of determination (R-squared) of a linear regression into nonnegative contributions from each explanatory variable using the Shapley value from cooperative game theory (Shapley, 1953, <doi:10.1515/9781400881970-018>). For each window, the exact Shapley value is computed by fitting all subsets of the explanatory variables and averaging the marginal contribution to R-squared across all orderings, which returns an order-invariant attribution that sums to the full-model R-squared. Use cases include variable importance, factor attribution, and feature selection in time-series regression. The package supports rolling and expanding windows, weights, and handling of missing values via min_obs', complete_obs', and na_restore arguments. The implementation uses the online and offline algorithms from the roll package to compute rolling and expanding cross-products efficiently with parallelism across columns and windows provided by RcppParallel'.

Estimation of the conditional covariance matrix using the RiskMetrics 2006 methodology of Zumbach (2007) <doi:10.2139/ssrn.1420185>.

Fits non-linear regression models on dependant data with Generalised Least Square (GLS) based Random Forest (RF-GLS) detailed in Saha, Basu and Datta (2021) <doi:10.1080/01621459.2021.1950003>.

Minimally adjust the values of numerical records in a data.frame, such that each record satisfies a predefined set of equality and/or inequality constraints. The constraints can be defined using the validate package. The core algorithms have recently been moved to the lintools package, refer to lintools for a more basic interface and access to a version of the algorithm that works with sparse matrices.

The concept of reliable and clinically significant change (Jacobson & Truax, 1991) helps you answer the following questions for a sample with two measurements at different points in time (pre & post): Which proportion of my sample has a (considering the reliability of the instrument) probably not-just-by-chance difference in pre- vs. post-scores? Which proportion of my sample does not only change in a statistically significant way (see question one), but also in a clinically significant way (e.g. change from a test score regarded "dysfunctional" to a score regarded "functional")? This package allows you to very easily create a scatterplot of your sample in which the x-axis maps to the pre-scores, the y-axis maps to the post-scores and several graphical elements (lines, colors) allow you to gain a quick overview about reliable changes in these scores. An example of this kind of plot is Figure 2 of Jacobson & Truax (1991). Referenced article: Jacobson, N. S., & Truax, P. (1991) <doi:10.1037/0022-006X.59.1.12>.

Calculates risk differences (or prevalence differences for cross-sectional data) and Number Needed to Treat (NNT) using generalized linear models with automatic link function selection. Provides robust model fitting with fallback methods, support for stratification and adjustment variables, inverse probability of treatment weighting (IPTW) for causal inference with NNT calculations, and publication-ready output formatting. Handles model convergence issues gracefully and provides confidence intervals using multiple approaches. Methods are based on approaches described in Mark W. Donoghoe and Ian C. Marschner (2018) "logbin: An R Package for Relative Risk Regression Using the Log-Binomial Model" <doi:10.18637/jss.v086.i09> for robust GLM fitting, Peter C. Austin (2011) "An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies" <doi:10.1080/00273171.2011.568786> for IPTW methods, and standard epidemiological methods for risk difference estimation as described in Kenneth J. Rothman, Sander Greenland and Timothy L. Lash (2008, ISBN:9780781755641) "Modern Epidemiology".

This package provides a robust alternative to the aJIVE (angle based Joint and Individual Variation Explained) method (Feng et al 2018: <doi:10.1016/j.jmva.2018.03.008>) for the estimation of joint and individual components in the presence of outliers in multi-source data. It decomposes the multi-source data into joint, individual and residual (noise) contributions. The decomposition is robust to outliers and noise in the data. The method is illustrated in Ponzi et al (2021) <arXiv:2101.09110>.

Assists in the manipulation and processing of linear features with the help of the sf package. Makes use of linear referencing to extract data from most shape files. Reference for this packages methods: Albeke, S.E. et al. (2010) <doi:10.1007/s10980-010-9528-4>.

Implementation of Gibbs sampling algorithm for Bayesian Estimation of the Reduced Reparameterized Unified Model ('rrum'), described by Culpepper and Hudson (2017) <doi: 10.1177/0146621617707511>.

Interface for multiple data sources, such as the `EDDS` API <https://evds2.tcmb.gov.tr/index.php?/evds/userDocs> of the Central Bank of the Republic of Türkiye and the `FRED` API <https://fred.stlouisfed.org/docs/api/fred/> of the Federal Reserve Bank. Both data providers require API keys for access, which users can easily obtain by creating accounts on their respective websites. The package provides caching ability with the selection of periods to increase the speed and efficiency of requests. It combines datasets requested from different sources, helping users when the data has common frequencies. While combining data frames whenever possible, it also keeps all requested data available as separate data frames to increase efficiency.

This package provides tools for analyzing stream networks, including graph construction, calculation of Link Magnitude, D-LINK, and export of spatial data. The Link Magnitude (Shreve stream order) method follows Shreve (1966) <doi:10.1086/627137>. The D-LINK metric follows Osborne & Wiley (1992) <doi:10.1139/f92-076>.

Collection of functions to evaluate sequences, decode hidden states and estimate parameters from a single or multiple sequences of a discrete time Hidden Markov Model. The observed values can be modeled by a multinomial distribution for categorical/labeled emissions, a mixture of Gaussians for continuous data and also a mixture of Poissons for discrete values. It includes functions for random initialization, simulation, backward or forward sequence evaluation, Viterbi or forward-backward decoding and parameter estimation using an Expectation-Maximization approach.

ROSE (RObust Semiparametric Efficient) random forests for robust semiparametric efficient estimation in partially parametric models (containing generalised partially linear models). Details can be found in the paper by Young and Shah (2024) <doi:10.48550/arXiv.2410.03471>.