Gradient boosting is a powerful statistical learning method known for its ability to model complex relationships between predictors and outcomes while performing inherent variable selection. However, traditional gradient boosting methods lack flexibility in handling longitudinal data where within-subject correlations play a critical role. In this package, we propose a novel approach Mixed Effect Gradient Boosting ('MEGB'), designed specifically for high-dimensional longitudinal data. MEGB incorporates a flexible semi-parametric model that embeds random effects within the gradient boosting framework, allowing it to account for within-individual covariance over time. Additionally, the method efficiently handles scenarios where the number of predictors greatly exceeds the number of observations (p>>n) making it particularly suitable for genomics data and other large-scale biomedical studies.

This package implements multi-factor curve analysis for grouped data in R', replicating and extending the functionality of the the Stata ado mfcurve (Krähmer, 2023) <https://ideas.repec.org/c/boc/bocode/s459224.html>. Related to the idea of specification curve analysis (Simonsohn, Simmons, and Nelson, 2020) <doi:10.1038/s41562-020-0912-z>. Includes data preprocessing, statistical testing, and visualization of results with confidence intervals.

Machine coded genetic algorithm (MCGA) is a fast tool for real-valued optimization problems. It uses the byte representation of variables rather than real-values. It performs the classical crossover operations (uniform) on these byte representations. Mutation operator is also similar to classical mutation operator, which is to say, it changes a randomly selected byte value of a chromosome by +1 or -1 with probability 1/2. In MCGAs there is no need for encoding-decoding process and the classical operators are directly applicable on real-values. It is fast and can handle a wide range of a search space with high precision. Using a 256-unary alphabet is the main disadvantage of this algorithm but a moderate size population is convenient for many problems. Package also includes multi_mcga function for multi objective optimization problems. This function sorts the chromosomes using their ranks calculated from the non-dominated sorting algorithm.

Several robust estimators for linear regression and variable selection are provided. Included are Maximum tangent likelihood estimator by Qin, et al., (2017), arXiv preprint <doi:10.48550/arXiv.1708.05439>, least absolute deviance estimator and Huber regression. The penalized version of each of these estimator incorporates L1 penalty function, i.e., LASSO and Adaptive Lasso. They are able to produce consistent estimates for both fixed and high-dimensional settings.

Parametric modeling of M-quantile regression coefficient functions.

This package contains functions to access movement data stored in movebank.org as well as tools to visualize and statistically analyze animal movement data, among others functions to calculate dynamic Brownian Bridge Movement Models. Move helps addressing movement ecology questions.

An implementation of a method for building simultaneous confidence intervals for the probabilities of a multinomial distribution given a set of observations, proposed by Sison and Glaz in their paper: Sison, C.P and J. Glaz. Simultaneous confidence intervals and sample size determination for multinomial proportions. Journal of the American Statistical Association, 90:366-369 (1995). The method is an R translation of the SAS code implemented by May and Johnson in their paper: May, W.L. and W.D. Johnson. Constructing two-sided simultaneous confidence intervals for multinomial proportions for small counts in a large number of cells. Journal of Statistical Software 5(6) (2000). Paper and code available at <DOI:10.18637/jss.v005.i06>.

When choosing proper variable selection methods, it is important to consider the uncertainty of a certain method. The model confidence bound for variable selection identifies two nested models (upper and lower confidence bound models) containing the true model at a given confidence level. A good variable selection method is the one of which the model confidence bound under a certain confidence level has the shortest width. When visualizing the variability of model selection and comparing different model selection procedures, model uncertainty curve is a good graphical tool. A good variable selection method is the one of whose model uncertainty curve will tend to arch towards the upper left corner. This function aims to obtain the model confidence bound and draw the model uncertainty curve of certain single model selection method under a coverage rate equal or little higher than user-given confidential level. About what model confidence bound is and how it work please see Li,Y., Luo,Y., Ferrari,D., Hu,X. and Qin,Y. (2019) Model Confidence Bounds for Variable Selection. Biometrics, 75:392-403. <DOI:10.1111/biom.13024>. Besides, flare is needed only you apply the SQRT or LAD method ('mcb totally has 8 methods). Although flare has been archived by CRAN, you can still get it in <https://CRAN.R-project.org/package=flare> and the latest version is useful for mcb'.

Allows the user to create graphs with multiple layers. The user can also modify the layers, the nodes, and the edges. The graph can also be visualized. Zaynab Hammoud and Frank Kramer (2018) <doi:10.3390/genes9110519>. More about multilayered graphs and their usage can be found in our review paper: Zaynab Hammoud and Frank Kramer (2020) <doi:10.1186/s41044-020-00046-0>.

Model fitting, sampling and visualization for the (Hidden) Markov Random Field model with pairwise interactions and general interaction structure from Freguglia, Garcia & Bicas (2020) <doi:10.1002/env.2613>, which has many popular models used in 2-dimensional lattices as particular cases, like the Ising Model and Potts Model. A complete manuscript describing the package is available in Freguglia & Garcia (2022) <doi:10.18637/jss.v101.i08>.

Developed for the following tasks. 1- simulating realizations from the canonical, restricted, and unrestricted finite mixture models. 2- Monte Carlo approximation for density function of the finite mixture models. 3- Monte Carlo approximation for the observed Fisher information matrix, asymptotic standard error, and the corresponding confidence intervals for parameters of the mixture models sing the method proposed by Basford et al. (1997) <https://espace.library.uq.edu.au/view/UQ:57525>.

The companion package provides all original data sets and functions that are used in the book "Model-Based Clustering and Classification for Data Science" by Charles Bouveyron, Gilles Celeux, T. Brendan Murphy and Adrian E. Raftery (2019, ISBN:9781108644181).

Extended tools for analyzing telemetry data using generalized hidden Markov models. Features of momentuHMM (pronounced ``momentum'') include data pre-processing and visualization, fitting HMMs to location and auxiliary biotelemetry or environmental data, biased and correlated random walk movement models, hierarchical HMMs, multiple imputation for incorporating location measurement error and missing data, user-specified design matrices and constraints for covariate modelling of parameters, random effects, decoding of the state process, visualization of fitted models, model checking and selection, and simulation. See McClintock and Michelot (2018) <doi:10.1111/2041-210X.12995>.

Data and code for the paper by Ehm, Gneiting, Jordan and Krueger ('Of Quantiles and Expectiles: Consistent Scoring Functions, Choquet Representations, and Forecast Rankings', JRSS-B, 2016 <DOI:10.1111/rssb.12154>).

Compute correlation and other association matrices from small to high-dimensional datasets with relative simple functions and sensible defaults. Includes options for shrinkage and robustness to improve results in noisy or high-dimensional settings (p >= n), plus convenient print/plot methods for inspection. Implemented with optimised C++ backends using BLAS/OpenMP and memory-aware symmetric updates. Works with base matrices and data frames, returning standard R objects via a consistent S3 interface. Useful across genomics, agriculture, and machine-learning workflows. Supports Pearson, Spearman, Kendall, distance correlation, partial correlation, and robust biweight mid-correlation; Blandâ Altman analyses and Lin's concordance correlation coefficient (including repeated-measures extensions). Methods based on Ledoit and Wolf (2004) <doi:10.1016/S0047-259X(03)00096-4>; Schäfer and Strimmer (2005) <doi:10.2202/1544-6115.1175>; Lin (1989) <doi:10.2307/2532051>.

This package contains model-based treatment of missing data for regression models with missing values in covariates or the dependent variable using maximum likelihood or Bayesian estimation (Ibrahim et al., 2005; <doi:10.1198/016214504000001844>; Luedtke, Robitzsch, & West, 2020a, 2020b; <doi:10.1080/00273171.2019.1640104><doi:10.1037/met0000233>). The regression model can be nonlinear (e.g., interaction effects, quadratic effects or B-spline functions). Multilevel models with missing data in predictors are available for Bayesian estimation. Substantive-model compatible multiple imputation can be also conducted.

This package provides functions to compute and plot multivariate (partial) Mantel correlograms.

This package implements methods for post-hoc analysis and visualisation of benchmark experiments, for mlr3 and beyond.

Finds the Maximum Likelihood (ML) Estimate of the mean vector and variance-covariance matrix for multivariate normal data with missing values.

We introduce a high-dimensional multi-study robust factor model, which learns latent features and accounts for the heterogeneity among source. It could be used for analyzing heterogeneous RNA sequencing data. More details can be referred to Jiang et al. (2025) <doi:10.48550/arXiv.2506.18478>.

Sometimes data for analysis are obtained using more convenient or less expensive means yielding "surrogate" variables for what could be obtained more accurately, albeit with less convenience; or less conveniently or at more expense yielding "reference" variables, thought of as being measured without error. Analysis of the surrogate variables measured with error generally yields biased estimates when the objective is to make inference about the reference variables. Often it is thought that ignoring the measurement error in surrogate variables only biases effects toward the null hypothesis, but this need not be the case. Measurement errors may bias parameter estimates either toward or away from the null hypothesis. If one has a data set with surrogate variable data from the full sample, and also reference variable data from a randomly selected subsample, then one can assess the bias introduced by measurement error in parameter estimation, and use this information to derive improved estimates based upon all available data. Formulaically these estimates based upon the reference variables from the validation subsample combined with the surrogate variables from the whole sample can be interpreted as starting with the estimate from reference variables in the validation subsample, and "augmenting" this with additional information from the surrogate variables. This suggests the term "augmented" estimate. The meerva package calculates these augmented estimates in the regression setting when there is a randomly selected subsample with both surrogate and reference variables. Measurement errors may be differential or non-differential, in any or all predictors (simultaneously) as well as outcome. The augmented estimates derive, in part, from the multivariate correlation between regression model parameter estimates from the reference variables and the surrogate variables, both from the validation subset. Because the validation subsample is chosen at random any biases imposed by measurement error, whether non-differential or differential, are reflected in this correlation and these correlations can be used to derive estimates for the reference variables using data from the whole sample. The main functions in the package are meerva.fit which calculates estimates for a dataset, and meerva.sim.block which simulates multiple datasets as described by the user, and analyzes these datasets, storing the regression coefficient estimates for inspection. The augmented estimates, as well as how measurement error may arise in practice, is described in more detail by Kremers WK (2021) <arXiv:2106.14063> and is an extension of the works by Chen Y-H, Chen H. (2000) <doi:10.1111/1467-9868.00243>, Chen Y-H. (2002) <doi:10.1111/1467-9868.00324>, Wang X, Wang Q (2015) <doi:10.1016/j.jmva.2015.05.017> and Tong J, Huang J, Chubak J, et al. (2020) <doi:10.1093/jamia/ocz180>.

This package provides functions to fit finite mixture of scale mixture of skew-normal (FM-SMSN) distributions, details in Prates, Lachos and Cabral (2013) <doi: 10.18637/jss.v054.i12>, Cabral, Lachos and Prates (2012) <doi:10.1016/j.csda.2011.06.026> and Basso, Lachos, Cabral and Ghosh (2010) <doi:10.1016/j.csda.2009.09.031>.

An implementation of the additive (Gurevitch et al., 2000 <doi:10.1086/303337>) and multiplicative (Lajeunesse, 2011 <doi:10.1890/11-0423.1>) factorial null models for multiple stressor data (Burgess et al., 2021 <doi:10.1101/2021.07.21.453207>). Effect sizes are able to be calculated for either null model, and subsequently classified into one of four different interaction classifications (e.g., antagonistic or synergistic interactions). Analyses can be conducted on data for single experiments through to large meta-analytical datasets. Minimal input (or statistical knowledge) is required, with any output easily understood. Summary figures are also able to be easily generated.

Integrates fairness auditing and bias mitigation methods for the mlr3 ecosystem. This includes fairness metrics, reporting tools, visualizations and bias mitigation techniques such as "Reweighing" described in Kamiran, Calders (2012) <doi:10.1007/s10115-011-0463-8> and "Equalized Odds" described in Hardt et al. (2016) <https://papers.nips.cc/paper/2016/file/9d2682367c3935defcb1f9e247a97c0d-Paper.pdf>. Integration with mlr3 allows for auditing of ML models as well as convenient joint tuning of machine learning algorithms and debiasing methods.