Create model-agnostic response-curve diagnostics for fitted prediction models. Supports profile curves, partial dependence, individual conditional expectation, and accumulated local effects; univariate curves, bivariate surfaces, ensemble summaries across multiple models, ALE-based interaction ranking, and optional raster-linked exploration with terra and shiny'. Static displays are returned as ggplot2 plots. For more details on the methods see Molnar (2025) <https://christophm.github.io/interpretable-ml-book/>.

In metabolic flux experiments tracer molecules (often glucose containing labelled carbon) are incorporated in compounds measured using mass spectrometry. The mass isotopologue distributions of these compounds needs to be corrected for natural abundance of labelled carbon and other effects, which are specific on the compound and ionization technique applied. This package provides functions to correct such effects in gas chromatography atmospheric pressure chemical ionization mass spectrometry analyses.

This package provides a lightweight delayed array abstraction for lazy, fused evaluation of multi-dimensional numeric data. Compared with Bioconductor's DelayedArray', it offers a small S3 API with tidy-friendly verbs, chunked materialisation, and optional backends for HDF5, memory mapping, and shared-memory parallelism. Expression trees are optimised before streaming evaluation, and results can be written directly to disk without fully materialising arrays in memory.

Decorrelates a set of summary statistics (i.e., Z-scores or P-values per SNP) via Decorrelation by Orthogonal Transformation (DOT) approach and performs gene-set analyses by combining transformed statistic values; operations are performed with algorithms that rely only on the association summary results and the linkage disequilibrium (LD). For more details on DOT and its power, see Olga (2020) <doi:10.1371/journal.pcbi.1007819>.

Replication methods to compute some basic statistic operations (means, standard deviations, frequency tables, percentiles, mean comparisons using weighted effect coding, generalized linear models, and linear multilevel models) in complex survey designs comprising multiple imputed or nested imputed variables and/or a clustered sampling structure which both deserve special procedures at least in estimating standard errors. See the package documentation for a more detailed description along with references.

Lightweight utilities to estimate autoregressive (AR) and autoregressive moving average (ARMA) noise models from residuals and apply matched generalized least squares to whiten functional magnetic resonance imaging (fMRI) design and data matrices. The ARMA estimator follows a classic 1982 approach <doi:10.1093/biomet/69.1.81>, and a restricted AR family mirrors workflows described by Cox (2012) <doi:10.1016/j.neuroimage.2011.08.056>.

Fair machine learning regression models which take sensitive attributes into account in model estimation. Currently implementing Komiyama et al. (2018) <http://proceedings.mlr.press/v80/komiyama18a/komiyama18a.pdf>, Zafar et al. (2019) <https://www.jmlr.org/papers/volume20/18-262/18-262.pdf> and my own approach from Scutari, Panero and Proissl (2022) <doi:10.1007/s11222-022-10143-w> that uses ridge regression to enforce fairness.

Fits a multivariate linear mixed effects model that uses a polygenic term, after Zhou & Stephens (2014) (<https://www.nature.com/articles/nmeth.2848>). Of particular interest is the estimation of variance components with restricted maximum likelihood (REML) methods. Genome-wide efficient mixed-model association (GEMMA), as implemented in the package gemma2', uses an expectation-maximization algorithm for variance components inference for use in quantitative trait locus studies.

This package contains a series of R functions for calculating various chance-corrected agreement coefficients (CAC) among 2 or more raters. Among the CAC coefficients covered are Cohen's kappa, Conger's kappa, Fleiss kappa, Brennan-Prediger coefficient, Gwet's AC1/AC2 coefficients, and Krippendorff's alpha. Multiple sets of weights are proposed for computing weighted analyses. Also included in this package is Bangdiwala's B coefficient.

Cleans, geocodes, validates, reviews, and exports messy location address data supplied by the user. The package sits on top of tidygeocoder': it calls geocoding services, rejects implausible coordinates with configurable region guards, applies fallback name/address matching, joins points to optional local geography with sf', and records an audit trail showing how each coordinate was produced. Outputs are designed for manual review, dashboards, and reusable location crosswalks.

This package provides a new method to implement clustering from multiple modality data of certain samples, the function M2SMjF() jointly factorizes multiple similarity matrices into a shared sub-matrix and several modality private sub-matrices, which is further used for clustering. Along with this method, we also provide function to calculate the similarity matrix and function to evaluate the best cluster number from the original data.

This package provides a collection of statistical tests for the detection of differential item functioning (DIF) in multistage tests. Methods entail logistic regression, an adaptation of the simultaneous item bias test (SIBTEST), and various score-based tests. The presented tests provide itemwise test for DIF along categorical, ordinal or metric covariates. Methods for uniform and non-uniform DIF effects are available depending on which method is used.

Compute effect sizes and their sampling variances from factorial experimental designs. The package supports calculation of simple effects, overall effects, and interaction effects for use in factorial meta-analyses. See Gurevitch et al. (2000) <doi:10.1086/303337>, Morris et al. (2007) <doi:10.1890/06-0442>, Lajeunesse (2011) <doi:10.1890/11-0423.1> and Macartney et al. (2022) <doi:10.1016/j.neubiorev.2022.104554>.

Builds, validates, analyzes, and visualizes network data from dyadic, event, matrix, igraph', and network inputs. Supports cross-sectional, longitudinal, bipartite, and multi-layer networks, with conversion helpers for common modeling workflows and plotting utilities for exploratory analysis. Network methods are described in Wasserman and Faust (1994, ISBN:9780521387071), Cranmer et al. (2021) <doi:10.1017/9781316662915>, and Minhas et al. (2022) <doi:10.1017/psrm.2021.56>.

Estimates DNA target concentration by classifying digital PCR (polymerase chain reaction) droplets as positive, negative, or rain, using Expectation-Maximization Clustering. The fitting is accomplished using the EMMIXskew R package (v. 1.0.3) by Kui Wang, Angus Ng, and Geoff McLachlan (2018) as based on their paper "Multivariate Skew t Mixture Models: Applications to Fluorescence-Activated Cell Sorting Data" <doi:10.1109/DICTA.2009.88>.

An R API providing access to a relational database with macroeconomic time series data for South Africa, obtained from the South African Reserve Bank (SARB) and Statistics South Africa (STATSSA), and updated on a weekly basis via the EconData <https://www.econdata.co.za/> platform and automated scraping of the SARB and STATSSA websites. The database is maintained at the Department of Economics at Stellenbosch University.

Estimation of group-based trajectory models, including finite mixture models for longitudinal data, supporting censored normal, zero-inflated Poisson, logit, and beta distributions, using expectation-maximization and quasi-Newton methods, with tools for model selection, diagnostics, and visualization of latent trajectory groups, <doi:10.4159/9780674041318>, Nagin, D. (2005). Group-Based Modeling of Development. Cambridge, MA: Harvard University Press. and Noel (2022), <https://orbilu.uni.lu/>, thesis.

Carries out analyses of two-way tables with one observation per cell, together with graphical displays for an additive fit and a diagnostic plot for removable non-additivity via a power transformation of the response. It implements methods from Tukey's Exploratory Data Analysis (1973) <ISBN: 978-0201076165>, including a 1-degree-of-freedom test for row*column non-additivity', linear in the row and column effects.

This package provides low-level access to GDAL functionality. GDAL is the Geospatial Data Abstraction Library a translator for raster and vector geospatial data formats that presents a single raster abstract data model and single vector abstract data model to the calling application for all supported formats <https://gdal.org/>. This package is focussed on providing exactly and only what GDAL does, to enable developing further tools.

This package provides methods for detecting structural breaks and estimating break locations for linear multiple regression models under general linear restrictions on the coefficient vector. Restrictions can be within regimes, across regimes, or both, and are supported in two forms: an affine parameterization (Form A: delta = S*theta + s) and explicit linear constraints (Form B: R*delta = r). Provides break date estimation with confidence intervals, a restricted sup-F test for the null of no structural change, simulation of critical values by Monte Carlo, and a bootstrap restart procedure to reduce the risk of convergence to spurious local optima. Also implements a generalized regression tree (linear model tree) procedure where each leaf contains a linear regression model rather than a local average. Reference: Perron, P., and Qu, Z. (2006). Estimating Restricted Structural Change Models. Journal of Econometrics, 134(2), 373-399. <doi:10.1016/j.jeconom.2005.06.030>.

This package provides a uniform interface to common response time preprocessing decisions that precede analysing aggregated response times or fitting an evidence accumulation model. Screening rules from different preprocessing routines -- absolute cutoffs, standard deviation and median absolute deviation criteria, recursive moving criteria, and model-based mixture flagging -- all return the same per-trial object, so that consequences of a preprocessing choice can be compared rather than assumed. The package also provides aggregation into EZ-diffusion summary statistics, diagnostics reporting what each rule removed and where rules disagree. Finally, generators for response time data with contaminants of known type are provided, so that a chosen pipeline can be tested against ground truth. Screening criteria follow Van Selst and Jolicoeur (1994) <doi:10.1080/14640749408401131>, the contaminant mixture Ratcliff and Tuerlinckx (2002) <doi:10.3758/BF03196302>, and the EZ-diffusion equations Wagenmakers, van der Maas and Grasman (2007) <doi:10.3758/BF03194023>.

This package is focused on finding differential exon usage using RNA-seq exon counts between samples with different experimental designs. It provides functions that allows the user to make the necessary statistical tests based on a model that uses the negative binomial distribution to estimate the variance between biological replicates and generalized linear models for testing. The package also provides functions for the visualization and exploration of the results.

AbSeq is a comprehensive bioinformatic pipeline for the analysis of sequencing datasets generated from antibody libraries and abseqR is one of its packages. AbseqR empowers the users of abseqPy with plotting and reporting capabilities and allows them to generate interactive HTML reports for the convenience of viewing and sharing with other researchers. Additionally, abseqR extends abseqPy to compare multiple repertoire analyses and perform further downstream analysis on its output.

This package includes functions and reference data to generate and manipulate log-ratios (also known as log size index (LSI) values) from measurements obtained on zooarchaeological material. Log ratios are used to compare the relative (rather than the absolute) dimensions of animals from archaeological contexts. The zoolog package is also able to seamlessly integrate data and references with heterogeneous nomenclature, which is internally managed by a zoolog thesaurus.