This package provides functions that (1) fit multivariate discrete distributions, (2) generate random numbers from multivariate discrete distributions, and (3) run regression and penalized regression on the multivariate categorical response data. Implemented models include: multinomial logit model, Dirichlet multinomial model, generalized Dirichlet multinomial model, and negative multinomial model. Making the best of the minorization-maximization (MM) algorithm and Newton-Raphson method, we derive and implement stable and efficient algorithms to find the maximum likelihood estimates. On a multi-core machine, multi-threading is supported.

The MARSS package provides maximum-likelihood parameter estimation for constrained and unconstrained linear multivariate autoregressive state-space (MARSS) models, including partially deterministic models. MARSS models are a class of dynamic linear model (DLM) and vector autoregressive model (VAR) model. Fitting available via Expectation-Maximization (EM), BFGS (using optim), and TMB (using the marssTMB companion package). Functions are provided for parametric and innovations bootstrapping, Kalman filtering and smoothing, model selection criteria including bootstrap AICb, confidences intervals via the Hessian approximation or bootstrapping, and all conditional residual types. See the user guide for examples of dynamic factor analysis, dynamic linear models, outlier and shock detection, and multivariate AR-p models. Online workshops (lectures, eBook, and computer labs) at <https://atsa-es.github.io/>.

This package implements an Integer Programming-based method for optimising genetic gain in polyclonal selection, where the goal is to select a group of genotypes that jointly meet multi-trait selection criteria. The method uses predictors of genotypic effects obtained from the fitting of mixed models. Its application is demonstrated with grapevine data, but is applicable to other species and breeding contexts. For more details see Surgy et al. (2025) <doi:10.1007/s00122-025-04885-0>.

We introduce factor models designed to jointly analyze high-dimensional count data from multiple studies by extracting study-shared and specified factors. Our factor models account for heterogeneous noises and overdispersion among counts with augmented covariates. We propose an efficient and speedy variational estimation procedure for estimating model parameters, along with a novel criterion for selecting the optimal number of factors and the rank of regression coefficient matrix. More details can be referred to Liu et al. (2024) <doi:10.48550/arXiv.2402.15071>.

An implementation of a Bayesian sparse group model using spike and slab priors in a regression context. It is designed for regression with a multivariate response variable, but also provides an implementation for univariate response.

Conduct random forests-based meta-analysis, obtain partial dependence plots for metaforest and classic meta-analyses, and cross-validate and tune metaforest- and classic meta-analyses in conjunction with the caret package. A requirement of classic meta-analysis is that the studies being aggregated are conceptually similar, and ideally, close replications. However, in many fields, there is substantial heterogeneity between studies on the same topic. Classic meta-analysis lacks the power to assess more than a handful of univariate moderators. MetaForest, by contrast, has substantial power to explore heterogeneity in meta-analysis. It can identify important moderators from a larger set of potential candidates (Van Lissa, 2020). This is an appealing quality, because many meta-analyses have small sample sizes. Moreover, MetaForest yields a measure of variable importance which can be used to identify important moderators, and offers partial prediction plots to explore the shape of the marginal relationship between moderators and effect size.

This package provides a generalised workflow for Matching-Adjusted Indirect Comparison (MAIC) analysis, which supports both anchored and non-anchored MAIC methods. In MAIC, unbiased trial outcome comparison is achieved by weighting the subject-level outcomes of the intervention trial so that the weighted aggregate measures of prognostic or effect-modifying variables match those of the comparator trial. Measurements supported include time-to-event (e.g., overall survival) and binary (e.g., objective tumor response). The method is described in Signorovitch et al. (2010) <doi:10.2165/11538370-000000000-00000> and Signorovitch et al. (2012) <doi:10.1016/j.jval.2012.05.004>.

Analyzes adverse events in clinical trials using the metalite data structure. The package simplifies the workflow to create production-ready tables, listings, and figures discussed in the adverse events analysis chapters of "R for Clinical Study Reports and Submission" by Zhang et al. (2022) <https://r4csr.org/>.

This package provides functions for MultiDimensional Feature Selection (MDFS): calculating multidimensional information gains, scoring variables, finding important variables, plotting selection results. This package includes an optional CUDA implementation that speeds up information gain calculation using NVIDIA GPGPUs. R. Piliszek et al. (2019) <doi:10.32614/RJ-2019-019>.

This package provides the biggest amount of statistical measures in the whole R world. Includes measures of regression, (multiclass) classification and multilabel classification. The measures come mainly from the mlr package and were programed by several mlr developers.

Perform sensitivity analysis on ordinary differential equation based models, including ad-hoc graphical analyses based on structured sequences of parameters as well as local sensitivity analysis. Functions are provided for creating inputs, simulating scenarios and plotting outputs.

This package implements state-of-the-art block bootstrap methods for extreme value statistics based on block maxima. Includes disjoint blocks, sliding blocks, relying on a circular transformation of blocks. Fast C++ backends (via Rcpp') ensure scalability for large time series.

Maximum likelihood estimates are obtained via an EM algorithm with either a first-order or a fully exponential Laplace approximation as documented by Broatch and Karl (2018) <doi:10.48550/arXiv.1710.05284>, Karl, Yang, and Lohr (2014) <doi:10.1016/j.csda.2013.11.019>, and by Karl (2012) <doi:10.1515/1559-0410.1471>. Karl and Zimmerman <doi:10.1016/j.jspi.2020.06.004> use this package to illustrate how the home field effect estimator from a mixed model can be biased under nonrandom scheduling.

Various reliability analysis methods for rare event inference (computing failure probability and quantile from model/function outputs).

Estimates monotone regression and variance functions in a nonparametric model, based on Dette, Holger, Neumeyer, and Pilz (2006) <doi:10.3150/bj/1151525131>.

Dataset and functions from the meta-analysis published in Medicine & Science in Sports & Exercise. It contains all the data and functions to reproduce the analysis. "Effectiveness of HIIE versus MICT in Improving Cardiometabolic Risk Factors in Health and Disease: A Meta-analysis". Felipe Mattioni Maturana, Peter Martus, Stephan Zipfel, Andreas M Nieà (2020) <doi:10.1249/MSS.0000000000002506>.

This package provides a tidy workflow for landscape-scale analysis. multilandr offers tools to generate landscapes at multiple spatial scales and compute landscape metrics, primarily using the landscapemetrics package. It also features utility functions for plotting and analyzing multi-scale landscapes, exploring correlations between metrics, filtering landscapes based on specific conditions, generating landscape gradients for a given metric, and preparing datasets for further statistical analysis. Documentation about multilandr is provided in an introductory vignette included in this package and in the paper by Huais (2024) <doi:10.1007/s10980-024-01930-z>; see citation("multilandr") for details.

This is a implementation of design methods for multi-state reliability demonstration tests (MSRDT) with failure count data, which is associated with the work from the published paper "Multi-state Reliability Demonstration Tests" by Suiyao Chen et al. (2017) <doi:10.1080/08982112.2017.1314493>. It implements two types of MSRDT, multiple periods (MP) and multiple failure modes (MFM). For MP, two different scenarios with criteria on cumulative periods (Cum) or separate periods (Sep) are implemented respectively. It also provides the implementation of conventional design method, namely binomial tests for failure count data.

Monte Carlo simulation is a stochastic method computing trajectories of photons in media. Surface backscattering is performing calculations in semi-infinite media and summarizing photon flux leaving the surface. This simulation is modeling the optical measurement of diffuse reflectance using an incident light beam. The semi-infinite media is considered to have flat surface. Media, typically biological tissue, is described by four optical parameters: absorption coefficient, scattering coefficient, anisotropy factor, refractive index. The media is assumed to be homogeneous. Computational parameters of the simulation include: number of photons, radius of incident light beam, lowest photon energy threshold, intensity profile (halo) radius, spatial resolution of intensity profile. You can find more information and validation in the Open Access paper. Laszlo Baranyai (2020) <doi:10.1016/j.mex.2020.100958>.

Multimodal mediation analysis is an emerging problem in microbiome data analysis. Multimedia make advanced mediation analysis techniques easy to use, ensuring that all statistical components are transparent and adaptable to specific problem contexts. The package provides a uniform interface to direct and indirect effect estimation, synthetic null hypothesis testing, bootstrap confidence interval construction, and sensitivity analysis. More details are available in Jiang et al. (2024) "multimedia: Multimodal Mediation Analysis of Microbiome Data" <doi:10.1101/2024.03.27.587024>.

Mixture model with overlapping clusters for binary actor-event data. Parameters are estimated in a Bayesian framework. Model and inference are described in Ranciati, Vinciotti, Wit (2017) Modelling actor-event network data via a mixture model under overlapping clusters. Submitted.

This package provides install functions of other languages such as java', python'.

This package provides a fast, flexible machine learning library, written in C++, that aims to provide fast, extensible implementations of cutting-edge machine learning algorithms. See also Curtin et al. (2023) <doi:10.21105/joss.05026>.

This package provides a simple tool allowing users to easily and dynamically explore or document a data set using a tree structure.