Distributes Gaussian process calculations across nodes in a distributed memory setting, using Rmpi. The bigGP class provides high-level methods for maximum likelihood with normal data, prediction, calculation of uncertainty (i.e., posterior covariance calculations), and simulation of realizations. In addition, bigGP provides an API for basic matrix calculations with distributed covariance matrices, including Cholesky decomposition, back/forwardsolve, crossproduct, and matrix multiplication.

Estimates Bayesian models of list experiments with informative priors. It includes functionalities to estimate different types of list experiment models with varying prior information. See Lu and Traunmüller (2026) <doi:10.1017/psrm.2025.10084> for examples and details of estimation.

Calculates several entropy metrics for spatial data inspired by Boltzmann's entropy formula. It includes metrics introduced by Cushman for landscape mosaics (Cushman (2015) <doi:10.1007/s10980-015-0305-2>), and landscape gradients and point patterns (Cushman (2021) <doi:10.3390/e23121616>); by Zhao and Zhang for landscape mosaics (Zhao and Zhang (2019) <doi:10.1007/s10980-019-00876-x>); and by Gao et al. for landscape gradients (Gao et al. (2018) <doi:10.1111/tgis.12315>; Gao and Li (2019) <doi:10.1007/s10980-019-00854-3>).

Package for Bayesian Variable Selection and Model Averaging in linear models and generalized linear models using stochastic or deterministic sampling without replacement from posterior distributions. Prior distributions on coefficients are from Zellner's g-prior or mixtures of g-priors corresponding to the Zellner-Siow Cauchy Priors or the mixture of g-priors from Liang et al (2008) <DOI:10.1198/016214507000001337> for linear models or mixtures of g-priors from Li and Clyde (2019) <DOI:10.1080/01621459.2018.1469992> in generalized linear models. Other model selection criteria include AIC, BIC and Empirical Bayes estimates of g. Sampling probabilities may be updated based on the sampled models using sampling w/out replacement or an efficient MCMC algorithm which samples models using a tree structure of the model space as an efficient hash table. See Clyde, Ghosh and Littman (2010) <DOI:10.1198/jcgs.2010.09049> for details on the sampling algorithms. Uniform priors over all models or beta-binomial prior distributions on model size are allowed, and for large p truncated priors on the model space may be used to enforce sampling models that are full rank. The user may force variables to always be included in addition to imposing constraints that higher order interactions are included only if their parents are included in the model. This material is based upon work supported by the National Science Foundation under Division of Mathematical Sciences grant 1106891. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

This package provides functions to fit, via Expectation-Maximization (EM) algorithm, the Bessel and Beta regressions to a data set with a bounded continuous response variable. The Bessel regression is a new and robust approach proposed in the literature. The EM version for the well known Beta regression is another major contribution of this package. See details in the references Barreto-Souza, Mayrink and Simas (2022) <doi:10.1111/anzs.12354> and Barreto-Souza, Mayrink and Simas (2020) <arXiv:2003.05157>.

This package implements efficient NumPy'-like broadcasted operations for atomic and recursive arrays. In the context of operations involving 2 (or more) arrays, â broadcastingâ (AKA singleton expansion) refers to efficiently recycling array dimensions, without making copies. Besides linking to Rcpp', broadcast does not use any external libraries in any way; broadcast was essentially made from scratch and can be installed out-of-the-box. The implementations available in broadcast include, but are not limited to, the following. 1) Broadcasted element-wise operations on any 2 arrays; they support a large set of relational, arithmetic, Boolean, string, and bit-wise operations. 2) A faster, more memory efficient, and broadcasted abind-like function, for binding arrays along an arbitrary dimension. 3) Broadcasted ifelse-like and apply-like functions. 4) Casting functions, that cast subset-groups of an array to a new dimension, cast nested lists to dimensional lists, and vice-versa. 5) A few linear algebra functions for statistics. The functions in the broadcast package strive to minimize computation time and memory usage (which is not just better for efficient computing, but also for the environment).

Instructor-developed tools for Analytics and Quantitative Methods (AQM) courses at Babson College. Included are compact descriptive statistics for data frames and lists, expanded reporting and graphics for linear regressions, and formatted reports for best subsets analyses.

This package provides functions streamlining the data analysis workflow: Outsourcing data import, renaming and type casting to a *.csv. Manipulating imputed datasets and fitting models on them. Summarizing models.

For a balanced design of experiments, this package calculates the sample size required to detect a certain standardized effect size, under a significance level. This package also provides three graphs; detectable standardized effect size vs power, sample size vs detectable standardized effect size, and sample size vs power, which show the mutual relationship between the sample size, power and the detectable standardized effect size. The detailed procedure is described in R. V. Lenth (2006-9) <https://homepage.divms.uiowa.edu/~rlenth/Power/>, Y. B. Lim (1998), M. A. Kastenbaum, D. G. Hoel and K. O. Bowman (1970) <doi:10.2307/2334851>, and Douglas C. Montgomery (2013, ISBN: 0849323312).

Build decision trees and random forests for classification and regression. The implementation strikes a balance between minimizing computing efforts and maximizing the expected predictive accuracy, thus scales well to large data sets. Multi-threading is available through OpenMP <https://gcc.gnu.org/wiki/openmp>.

Approximate Euclidean k-nearest neighbour search routines that operate on bigmemory::big.matrix data through Annoy indexes created with RcppAnnoy'. The package builds persistent on-disk indexes plus sidecar metadata from streamed big.matrix rows, supports euclidean, angular, Manhattan, and dot-product Annoy metrics, and can either return in-memory results or stream neighbour indices and distances into destination bigmemory matrices. Explicit index life cycle helpers, stronger metadata validation, descriptor-aware file-backed workflows, and benchmark helpers are also included.

Create randomizations for block random clinical trials. Can also produce a pdf file of randomization cards.

Bayesian analysis for stochastic extensions of non-linear dynamic systems using advanced computational algorithms. Described in Bouranis, L., Demiris, N., Kalogeropoulos, K., and Ntzoufras, I. (2022) <doi:10.48550/arXiv.2211.15229>.

This package provides functions to download and tidy statistical data published by the Bank of England <https://www.bankofengland.co.uk>. Covers Bank Rate, SONIA', gilt yields, exchange rates, mortgage rates, mortgage approvals, consumer credit, and money supply. Series are fetched from the Bank of England Interactive Statistical Database using its CSV endpoint. Data is cached locally between sessions.

This package provides branding, theme application, and navigation utilities for applications built with bs4Dash and shiny'. Supports configurable sidebar brand display modes, hover-expand behavior, and theme customization using CSS variables. Includes standardized navigation components such as refresh and help controls, along with helpers for common navigation bar and footer layouts.

MCMC algorithms & processing functions for: 1. single response multiple regression, see Papageorgiou, G. (2018) <doi: 10.32614/RJ-2018-069>, 2. multivariate response multiple regression, with nonparametric models for the means, the variances and the correlation matrix, with variable selection, see Papageorgiou, G. and Marshall, B. C. (2020) <doi: 10.1080/10618600.2020.1739534>, 3. joint mean-covariance models for multivariate responses, see Papageorgiou, G. (2022) <doi: 10.1002/sim.9376>, and 4.Dirichlet process mixtures, see Papageorgiou, G. (2019) <doi: 10.1111/anzs.12273>.

Maximum likelihood estimation of copula-based zero-inflated (and non-inflated) Poisson and negative binomial count models, based on the article <doi:10.18637/jss.v109.i01>. Supports Frank and Gaussian copulas. Allows for mixed margins (e.g., one margin Poisson, the other zero-inflated negative binomial), and several marginal link functions. Built-in methods for publication-quality tables using texreg', post-estimation diagnostics using DHARMa', and testing for marginal zero-modification via <doi:10.1177/0962280217749991>. For information on copula regression for count data, see Genest and Nešlehová (2007) <doi:10.1017/S0515036100014963> as well as Nikoloulopoulos (2013) <doi:10.1007/978-3-642-35407-6_11>. For information on zero-inflated count regression generally, see Lambert (1992) <https://www.jstor.org/stable/1269547>. The author acknowledges support by NSF DMS-1925119 and DMS-212324.

The R-package bayespm implements Bayesian Statistical Process Control and Monitoring (SPC/M) methodology. These methods utilize available prior information and/or historical data, providing efficient online quality monitoring of a process, in terms of identifying moderate/large transient shifts (i.e., outliers) or persistent shifts of medium/small size in the process. These self-starting, sequentially updated tools can also run under complete absence of any prior information. The Predictive Control Charts (PCC) are introduced for the quality monitoring of data from any discrete or continuous distribution that is a member of the regular exponential family. The Predictive Ratio CUSUMs (PRC) are introduced for the Binomial, Poisson and Normal data (a later version of the library will cover all the remaining distributions from the regular exponential family). The PCC targets transient process shifts of typically large size (a.k.a. outliers), while PRC is focused in detecting persistent (structural) shifts that might be of medium or even small size. Apart from monitoring, both PCC and PRC provide the sequentially updated posterior inference for the monitored parameter. Bourazas K., Kiagias D. and Tsiamyrtzis P. (2022) "Predictive Control Charts (PCC): A Bayesian approach in online monitoring of short runs" <doi:10.1080/00224065.2021.1916413>, Bourazas K., Sobas F. and Tsiamyrtzis, P. 2023. "Predictive ratio CUSUM (PRC): A Bayesian approach in online change point detection of short runs" <doi:10.1080/00224065.2022.2161434>, Bourazas K., Sobas F. and Tsiamyrtzis, P. 2023. "Design and properties of the predictive ratio cusum (PRC) control charts" <doi:10.1080/00224065.2022.2161435>.

Perform bootstrap-based hypothesis testing procedures on three statistical problems. In particular, it covers independence testing, testing the slope in a linear regression setting, and goodness-of-fit testing, following (Derumigny, Galanis, Schipper and Van der Vaart, 2025) <doi:10.48550/arXiv.2512.10546>.

This package provides a framework for building interactive dashboards and document-based reports. Underlying data manipulation and visualization is possible using a web-based point and click user interface.

Fully Bayesian Classification with a subset of high-dimensional features, such as expression levels of genes. The data are modeled with a hierarchical Bayesian models using heavy-tailed t distributions as priors. When a large number of features are available, one may like to select only a subset of features to use, typically those features strongly correlated with the response in training cases. Such a feature selection procedure is however invalid since the relationship between the response and the features has be exaggerated by feature selection. This package provides a way to avoid this bias and yield better-calibrated predictions for future cases when one uses F-statistic to select features.

Managing and generating standardised text for methods and results sections of scientific reports. It handles template variable substitution and supports hierarchical organisation of text through dot-separated paths. The package supports both RDS and JSON database formats, enabling version control and cross-language compatibility.

Calculates the necessary quantities to perform Bayesian multigroup equivalence testing. Currently the package includes the Bayesian models and equivalence criteria outlined in Pourmohamad and Lee (2023) <doi:10.1002/sta4.645>, but more models and equivalence testing features may be added over time.

This package provides tools and code snippets for summarizing nested data, adverse events and REDCap study information.