This package provides a framework for scalable statistical computing on large on-disk matrices stored in HDF5 files. It provides efficient block-wise implementations of core linear-algebra operations (matrix multiplication, SVD, PCA, QR decomposition, and canonical correlation analysis) written in C++ and R. These building blocks are designed not only for direct use, but also as foundational components for developing new statistical methods that must operate on datasets too large to fit in memory. The package supports data provided either as HDF5 files or standard R objects, and is intended for high-dimensional applications such as omics and precision-medicine research.

This package implements three test procedures using bootstrap resampling techniques for assessing treatment effects in one-way ANOVA models with unequal variances (heteroscedasticity). It includes a parametric bootstrap likelihood ratio test (PB_LRT()), a pairwise parametric bootstrap mean test (PPBMT()), and a Rademacher wild pairwise non-parametric bootstrap test (RWPNPBT()). These methods provide robust alternatives to classical ANOVA and standard pairwise comparisons when the assumption of homogeneity of variances is violated.

Gibbs sampling for Bayesian spatial blind source separation (BSP-BSS). BSP-BSS is designed for spatially dependent signals in high dimensional and large-scale data, such as neuroimaging. The method assumes the expectation of the observed images as a linear mixture of multiple sparse and piece-wise smooth latent source signals, and constructs a Bayesian nonparametric prior by thresholding Gaussian processes. Details can be found in our paper: Wu, B., Guo, Y., & Kang, J. (2024). Bayesian spatial blind source separation via the thresholded gaussian process. Journal of the American Statistical Association, 119(545), 422-433.

Currently, the package provides several functions for plotting and analyzing bibliometric data (JIF, Journal Impact Factor, and paper percentile values), beamplots with citations and percentiles, and three plot functions to visualize the result of a reference publication year spectroscopy (RPYS) analysis performed in the free software CRExplorer (see <http://crexplorer.net>). Further extension to more plot variants is planned.

Twelve confidence intervals for one binomial proportion or a vector of binomial proportions are computed. The confidence intervals are: Jeffreys, Wald, Wald corrected, Wald, Blyth and Still, Agresti and Coull, Wilson, Score, Score corrected, Wald logit, Wald logit corrected, Arcsine and Exact binomial. References include, among others: Vollset, S. E. (1993). "Confidence intervals for a binomial proportion". Statistics in Medicine, 12(9): 809-824. <doi:10.1002/sim.4780120902>.

This package provides comprehensive tools for blinded sample size re-estimation (BSSR) in two-arm clinical trials with binary endpoints. Unlike traditional fixed-sample designs, BSSR allows adaptive sample size adjustments during trials while maintaining statistical integrity and study blinding. Implements five exact statistical tests: Pearson chi-squared, Fisher exact, Fisher mid-p, Z-pooled exact unconditional, and Boschloo exact unconditional tests. Supports restricted, unrestricted, and weighted BSSR approaches with exact Type I error control. Statistical methods based on Mehrotra et al. (2003) <doi:10.1111/1541-0420.00051> and Kieser (2020) <doi:10.1007/978-3-030-49528-2_21>.

Perform mediation analysis in the presence of high-dimensional mediators based on the potential outcome framework. Bayesian Mediation Analysis (BAMA), developed by Song et al (2019) <doi:10.1111/biom.13189> and Song et al (2020) <doi:10.48550/arXiv.2009.11409>, relies on two Bayesian sparse linear mixed models to simultaneously analyze a relatively large number of mediators for a continuous exposure and outcome assuming a small number of mediators are truly active. This sparsity assumption also allows the extension of univariate mediator analysis by casting the identification of active mediators as a variable selection problem and applying Bayesian methods with continuous shrinkage priors on the effects.

This package creates an area-proportional Venn diagram of 2 or 3 circles. BioVenn is the only R package that can automatically generate an accurate area-proportional Venn diagram by having only lists of (biological) identifiers as input. Also offers the option to map Entrez and/or Affymetrix IDs to Ensembl IDs. In SVG mode, text and numbers can be dragged and dropped. Based on the BioVenn web interface available at <https://www.biovenn.nl>. Hulsen (2021) <doi:10.3233/DS-210032>.

The Biodem package provides a number of functions for Biodemographic analysis.

This package provides a suite of Bayesian MI-LASSO for variable selection methods for multiply-imputed datasets. The package includes four Bayesian MI-LASSO models using shrinkage (Multi-Laplace, Horseshoe, ARD) and Spike-and-Slab (Spike-and-Laplace) priors, along with tools for model fitting via MCMC, four-step projection predictive variable selection, and hyperparameter calibration. Methods are suitable for both continuous and binary covariates under missing-at-random or missing-completely-at-random assumptions. See Zou, J., Wang, S. and Chen, Q. (2025), Bayesian MI-LASSO for Variable Selection on Multiply-Imputed Data. ArXiv, 2211.00114. <doi:10.48550/arXiv.2211.00114> for more details. We also provide the frequentist`s MI-LASSO function.

Bayesian inference on the (discrete) power spectrum of time series.

Compare dissolution profiles with confidence interval of similarity factor f2 using bootstrap methodology as described in the literature, such as Efron and Tibshirani (1993, ISBN:9780412042317), Davison and Hinkley (1997, ISBN:9780521573917), and Shah et al. (1998) <doi:10.1023/A:1011976615750>. The package can also be used to simulate dissolution profiles based on mathematical modelling and multivariate normal distribution.

This package provides a framework and toolkit to guide R dashboard developers in implementing the Behavioral Insight Design (BID) framework. The package offers functions for documenting each of the five stages (Interpret, Notice, Anticipate, Structure, and Validate), along with a comprehensive concept dictionary. Works with both shiny applications and Quarto dashboards.

Enable users to evaluate long-term trends using a Generalized Additive Modeling (GAM) approach. The model development includes selecting a GAM structure to describe nonlinear seasonally-varying changes over time, incorporation of hydrologic variability via either a river flow or salinity, the use of an intervention to deal with method or laboratory changes suspected to impact data values, and representation of left- and interval-censored data. The approach has been applied to water quality data in the Chesapeake Bay, a major estuary on the east coast of the United States to provide insights to a range of management- and research-focused questions. Methodology described in Murphy (2019) <doi:10.1016/j.envsoft.2019.03.027>.

This package provides a toolbox for analyzing and simulating large networks based on hierarchical exponential-family random graph models (HERGMs).'bigergm implements the estimation for large networks efficiently building on the lighthergm and hergm packages. Moreover, the package contains tools for simulating networks with local dependence to assess the goodness-of-fit.

Running and comparing meta-analyses of data with hierarchical Bayesian models in Stan, including convenience functions for formatting data, plotting and pooling measures specific to meta-analysis. This implements many models from Meager (2019) <doi:10.1257/app.20170299>.

Network meta-analyses using Bayesian framework following Dias et al. (2013) <DOI:10.1177/0272989X12458724>. Based on the data input, creates prior, model file, and initial values needed to run models in rjags'. Able to handle binomial, normal and multinomial arm-level data. Can handle multi-arm trials and includes methods to incorporate covariate and baseline risk effects. Includes standard diagnostics and visualization tools to evaluate the results.

This package provides a tool to perform all different statistical tests and calculations needed by Biological dosimetry Laboratories. Detailed documentation is available in <https://biodosetools-team.github.io/documentation/>.

This package performs Bayesian variable screening and selection for ultra-high dimensional linear regression models.

Fits novel models for the conditional relative risk, risk difference and odds ratio <doi:10.1080/01621459.2016.1192546>.

Randomly reassigns the group identifications to one of the variables of the database, say Treatment, and randomly reassigns the observation numbers of the dataset. Reorders the observations according to these new numbers. Centers each group of Treatment at the grand mean in order to further mask the treatment. An unmasking function is provided so that the user can identify the potential outliers in terms of their original values when blinding is no longer needed. It is suggested that a forward search procedure be performed on the masked data. Details of some forward search functions may be found in <https://CRAN.R-project.org/package=forsearch>.

Statistical classification and regression have been popular among various fields and stayed in the limelight of scientists of those fields. Examples of the fields include clinical trials where the statistical classification of patients is indispensable to predict the clinical courses of diseases. Considering the negative impact of diseases on performing daily tasks, correctly classifying patients based on the clinical information is vital in that we need to identify patients of the high-risk group to develop a severe state and arrange medical treatment for them at an opportune moment. Deep learning - a part of artificial intelligence - has gained much attention, and research on it burgeons during past decades: see, e.g, Kazemi and Mirroshandel (2018) <DOI:10.1016/j.artmed.2017.12.001>. It is a veritable technique which was originally designed for the classification, and hence, the Buddle package can provide sublime solutions to various challenging classification and regression problems encountered in the clinical trials. The Buddle package is based on the back-propagation algorithm - together with various powerful techniques such as batch normalization and dropout - which performs a multi-layer feed-forward neural network: see Krizhevsky et. al (2017) <DOI:10.1145/3065386>, Schmidhuber (2015) <DOI:10.1016/j.neunet.2014.09.003> and LeCun et al. (1998) <DOI:10.1109/5.726791> for more details. This package contains two main functions: TrainBuddle() and FetchBuddle(). TrainBuddle() builds a feed-forward neural network model and trains the model. FetchBuddle() recalls the trained model which is the output of TrainBuddle(), classifies or regresses given data, and make a final prediction for the data.

The Bayesian Adjustment for Confounding (BAC) algorithm (Wang et al., 2012) can be used to estimate the causal effect of a continuous exposure on a continuous outcome. This package provides an approximate sensitivity analysis of BAC with regards to the hyperparameter omega. BACprior also provides functions to guide the user in their choice of an appropriate omega value. The method is based on Lefebvre, Atherton and Talbot (2014).

Datasets and functions for the book "Initiation à la Statistique avec R", F. Bertrand and M. Maumy-Bertrand (2022, ISBN:978-2100782826 Dunod, fourth edition).