This package provides a modeltime extension that implements time series ensemble forecasting methods including model averaging, weighted averaging, and stacking. These techniques are popular methods to improve forecast accuracy and stability.

This package provides tools to create a layout for figures made of multiple panels, and to fill the panels with base, lattice', ggplot2 and ComplexHeatmap plots, grobs, as well as content from all image formats supported by ImageMagick (accessed through magick').

The main objective of this package is to support the definition of Moodle elements taking advantage of the power that R offers. In this first version, it allows the definition of quizzes to be included in the question bank.

This package contains the Markov cluster algorithm (MCL) for identifying clusters in networks and graphs. The algorithm simulates random walks on a (n x n) matrix as the adjacency matrix of a graph. It alternates an expansion step and an inflation step until an equilibrium state is reached.

Bending non-positive-definite (symmetric) matrices to positive-definite, using weighted and unweighted methods. Jorjani, H., et al. (2003) <doi:10.3168/jds.S0022-0302(03)73646-7>. Schaeffer, L. R. (2014) <http://animalbiosciences.uoguelph.ca/~lrs/ELARES/PDforce.pdf>.

Compute the coefficient of determination for outcomes in n-dimensions. May be useful for multidimensional predictions (such as a multinomial model) or calculating goodness of fit from latent variable models such as probabilistic topic models like latent Dirichlet allocation or deterministic topic models like latent semantic analysis. Based on Jones (2019) <arXiv:1911.11061>.

Multi-criteria design of experiments algorithm that simultaneously optimizes up to six different criteria ('I', Id', D', Ds', A and As'). The algorithm finds the optimal Pareto front and, if requested, selects a possible symmetrical design on it. The symmetrical design is selected based on two techniques: minimum distance with the Utopia point or the TOPSIS approach.

Multiscale moving sum procedure for the detection of changes in expectation in univariate sequences. References - Multiscale change point detection via gradual bandwidth adjustment in moving sum processes (2021+), Tijana Levajkovic and Michael Messer.

Subset a control group to match an intervention group on a set of features using multivariate matching and propensity score calipers. Based on methods in Rosenbaum and Rubin (1985).

This package performs Modal Clustering (MAC) including Hierarchical Modal Clustering (HMAC) along with their parallel implementation (PHMAC) over several processors. These model-based non-parametric clustering techniques can extract clusters in very high dimensions with arbitrary density shapes. By default clustering is performed over several resolutions and the results are summarised as a hierarchical tree. Associated plot functions are also provided. There is a package vignette that provides many examples. This version adheres to CRAN policy of not spanning more than two child processes by default.

This package provides a tool for implementing so called deft approach (see Fisher, David J., et al. (2017) <DOI:10.1136/bmj.j573>) and model visualization.

This package provides methods for estimating and utilizing the multivariate generalized propensity score (mvGPS) for multiple continuous exposures described in Williams, J.R, and Crespi, C.M. (2020) <arxiv:2008.13767>. The methods allow estimation of a dose-response surface relating the joint distribution of multiple continuous exposure variables to an outcome. Weights are constructed assuming a multivariate normal density for the marginal and conditional distribution of exposures given a set of confounders. Confounders can be different for different exposure variables. The weights are designed to achieve balance across all exposure dimensions and can be used to estimate dose-response surfaces.

The current version of the MixSAL package allows users to generate data from a multivariate SAL distribution or a mixture of multivariate SAL distributions, evaluate the probability density function of a multivariate SAL distribution or a mixture of multivariate SAL distributions, and fit a mixture of multivariate SAL distributions using the Expectation-Maximization (EM) algorithm (see Franczak et. al, 2014, <doi:10.1109/TPAMI.2013.216>, for details).

Weakly supervised (WS), multiple instance (MI) data lives in numerous interesting applications such as drug discovery, object detection, and tumor prediction on whole slide images. The mildsvm package provides an easy way to learn from this data by training Support Vector Machine (SVM)-based classifiers. It also contains helpful functions for building and printing multiple instance data frames. The core methods from mildsvm come from the following references: Kent and Yu (2024) <doi:10.1214/24-AOAS1876>; Xiao, Liu, and Hao (2018) <doi:10.1109/TNNLS.2017.2766164>; Muandet et al. (2012) <https://proceedings.neurips.cc/paper/2012/file/9bf31c7ff062936a96d3c8bd1f8f2ff3-Paper.pdf>; Chu and Keerthi (2007) <doi:10.1162/neco.2007.19.3.792>; and Andrews et al. (2003) <https://papers.nips.cc/paper/2232-support-vector-machines-for-multiple-instance-learning.pdf>. Many functions use the Gurobi optimization back-end to improve the optimization problem speed; the gurobi R package and associated software can be downloaded from <https://www.gurobi.com> after obtaining a license.

This package contains a suite of functions for health economic evaluations with missing outcome data. The package can fit different types of statistical models under a fully Bayesian approach using the software JAGS (which should be installed locally and which is loaded in missingHE via the R package R2jags'). Three classes of models can be fitted under a variety of missing data assumptions: selection models, pattern mixture models and hurdle models. In addition to model fitting, missingHE provides a set of specialised functions to assess model convergence and fit, and to summarise the statistical and economic results using different types of measures and graphs. The methods implemented are described in Mason (2018) <doi:10.1002/hec.3793>, Molenberghs (2000) <doi:10.1007/978-1-4419-0300-6_18> and Gabrio (2019) <doi:10.1002/sim.8045>.

This package provides functions for estimating structural equation models using instrumental variables.

Generates mid upper arm circumference (MUAC) and body mass index (BMI) for age z-scores and percentiles based on LMS method for children and adolescents up to 19 years that can be used to assess nutritional and health status and define risk of adverse health events.

Simplifies Brazilian names phonetically using a custom metaphoneBR algorithm that preserves ending vowels. Useful for name matching processing preserving gender information carried generally by ending vowels in Portuguese. Mation (2025) <doi:10.6082/uchicago.15104>.

Multi-omic (or any multi-view) spectral clustering methods often assume the same number of clusters across all datasets. We supply methods for multi-omic spectral clustering when the number of distinct clusters differs among the omics profiles (views).

Estimation, inference and diagnostics for Univariate Autoregressive Markov Switching Models for Linear and Generalized Models. Distributions for the series include gaussian, Poisson, binomial and gamma cases. The EM algorithm is used for estimation (see Perlin (2012) <doi:10.2139/ssrn.1714016>).

An implementation of the expectation conditional maximization (ECM) algorithm for matrix-variate variance gamma (MVVG) and normal-inverse Gaussian (MVNIG) linear models. These models are designed for settings of multivariate analysis with clustered non-uniform observations and correlated responses. The package includes fitting and prediction functions for both models, and an example dataset from a periodontal on Gullah-speaking African Americans, with responses in gaad_res', and covariates in gaad_cov'. For more details on the matrix-variate distributions used, see Gallaugher & McNicholas (2019) <doi:10.1016/j.spl.2018.08.012>.

This package provides a nature-inspired metaheuristic algorithm based on the echolocation behavior of microbats that uses frequency tuning to optimize problems in both continuous and discrete dimensions. This R package makes it easy to implement the standard bat algorithm on any user-supplied function. The algorithm was first developed by Xin-She Yang in 2010 (<DOI:10.1007/978-3-642-12538-6_6>, <DOI:10.1109/CINTI.2014.7028669>).

This package provides a set of functions, classes and methods for performing ABC and ABC/XYZ analyses, identifying overperforming, underperforming and constantly performing items, and plotting, analyzing as well as predicting the temporal development of items.

This package implements a regularized Bayesian estimator that optimizes the estimation of between-group coefficients for multilevel latent variable models by minimizing mean squared error (MSE) and balancing variance and bias. The package provides more reliable estimates in scenarios with limited data, offering a robust solution for accurate parameter estimation in two-level latent variable models. It is designed for researchers in psychology, education, and related fields who face challenges in estimating between-group effects under small sample sizes and low intraclass correlation coefficients. The package includes comprehensive S3 methods for result objects: print(), summary(), coef(), se(), vcov(), confint(), as.data.frame(), dim(), length(), names(), and update() for enhanced usability and integration with standard R workflows. Dashuk et al. (2025a) <doi:10.1017/psy.2025.10045> derived the optimal regularized Bayesian estimator; Dashuk et al. (2025b) <doi:10.1007/s41237-025-00264-7> extended it to the multivariate case; and Luedtke et al. (2008) <doi:10.1037/a0012869> formalized the two-level latent variable framework.