This package provides a wrapper of different standard estimation methods for gravity models. This package provides estimation methods for log-log models and multiplicative models.

This package provides a collection of fast k-means clustering algorithms under a single, uniform interface. The core method is Geometric-k-means, a bound-free algorithm of Sharma et al. (2026) <doi:10.1007/s10994-025-06891-1> that uses geometry to restrict computation to the data points able to change clusters, substantially reducing distance computations and runtime while returning the same result as standard k-means. Also included are Lloyd's algorithm, Elkan, Hamerly, Annulus, Exponion, and Ball k-means. All algorithms are implemented in C++ via Rcpp and RcppEigen and return the final centroids, optional per-point cluster assignments, and computational statistics.

Implementation of several goodness-of-fit tests for functional data. Currently, mostly related with the functional linear model with functional/scalar response and functional/scalar predictor. The package allows for the replication of the data applications considered in Garcà a-Portugués, à lvarez-Liébana, à lvarez-Pérez and González-Manteiga (2021) <doi:10.1111/sjos.12486>.

Simulates from discrete and continuous target distributions using geometric Metropolis-Hastings (MH) algorithms. Users specify the target distribution by an R function that evaluates the log un-normalized pdf or pmf. The package also contains a function implementing a specific geometric MH algorithm for performing high-dimensional Bayesian variable selection.

An implementation of functions to display Greek letters on the RStudio (include subscript and superscript indexes) and RGui (without subscripts and only with superscript 1, 2 or 3; because RGui doesn't support printing the corresponding Unicode characters as a string: all subscripts ranging from 0 to 9 and superscripts equal to 0, 4, 5, 6, 7, 8 or 9). The functions in this package do not work properly on the R console. Characters are used via Unicode and encoded as UTF-8 to ensure that they can be viewed on all operating systems. Other characters related to mathematics are included, such as the infinity symbol. All this accessible from very simple commands. This is a package that can be used for teaching purposes, the statistical notation for hypothesis testing can be written from this package and so it is possible to build a course from the swirlify package. Another utility of this package is to create new summary functions that contain the functional form of the model adjusted with the Greek letters, thus making the transition from statistical theory to practice easier. In addition, it is a natural extension of the clisymbols package.

Turns one-off iterative R procedures (such as for loops, lapply() or pmap() from purrr') into production-grade workflows by wrapping them with orthogonal, composable execution layers. Two layers are always active: structured logging with real traceback and per-case timing; and reproducibility capture, which records the R version, loaded package versions, execution environment, the exact iteration mask, and a stat-based fingerprint of every input file referenced in the mask (with a diff_inputs() helper to detect silent drift between runs). Parallel execution (built on the future framework, Bengtsson (2021) <doi:10.32614/RJ-2021-048>), non-blocking background jobs, and opt-in progress reporting (via progressr') are implemented as optional, composable layers. Further layers (error replay, content-hash input fingerprinting, content-based case identifiers) are planned and will remain composable with the default layers.

This package implements the five-parameter Generalized Kumaraswamy ('gkw') distribution proposed by Carrasco, Ferrari and Cordeiro (2010) <doi:10.48550/arXiv.1004.0911> and its seven nested sub-families for modeling bounded continuous data on the unit interval (0,1). The gkw distribution extends the Kumaraswamy distribution described by Jones (2009) <doi:10.1016/j.stamet.2008.04.001>. Provides density, distribution, quantile, and random generation functions, along with analytical log-likelihood, gradient, and Hessian functions implemented in C++ via RcppArmadillo for maximum computational efficiency. Suitable for modeling proportions, rates, percentages, and indices exhibiting complex features such as asymmetry, or heavy tails and other shapes not adequately captured by standard distributions like simple Beta or Kumaraswamy.

Implementation of various inference and simulation tools to apply generalized additive models to bivariate dependence structures and non-simplified vine copulas.

Implementation of a Bayesian approach for estimating a mixture of gamma distributions in which the mixing occurs over the shape parameter. This family provides a flexible and novel approach for modeling heavy-tailed distributions, it is computationally efficient, and it only requires to specify a prior distribution for a single parameter.

This package provides a genetic algorithm framework for regression problems requiring discrete optimization over model spaces with unknown or varying dimension, where gradient-based methods and exhaustive enumeration are impractical. Uses a compact chromosome representation for tasks including spline knot placement and best-subset variable selection, with constraint-preserving crossover and mutation, exact uniform initialization under spacing constraints, steady-state replacement, and optional island-model parallelization from Lu, Lund, and Lee (2010, <doi:10.1214/09-AOAS289>). The computation is built on the GA engine of Scrucca (2017, <doi:10.32614/RJ-2017-008>) and changepointGA engine from Li and Lu (2024, <doi:10.48550/arXiv.2410.15571>). In challenging high-dimensional settings, GAReg enables efficient search and delivers near-optimal solutions when alternative algorithms are not well-justified.

Finite candidate-set approximate optimal designs for group testing and related experiments, using convex optimization and equivalence checks. Implements the information matrix and cost structure for the prevalence / sensitivity / specificity model used in Huang and colleagues (2020), as in Chi-Kuang Yeh, Weng Kee Wong, and Julie Zhou (<doi:10.48550/arXiv.2508.08445>).

This package provides a genetic algorithm for finding variable subsets in high dimensional data with high prediction performance. The genetic algorithm can use ordinary least squares (OLS) regression models or partial least squares (PLS) regression models to evaluate the prediction power of variable subsets. By supporting different cross-validation schemes, the user can fine-tune the tradeoff between speed and quality of the solution.

Simple tools to draw sky maps in ggplot2 using galactic or equatorial coordinates. Includes custom coordinate systems, grid labels, and helpers for sky map breaks.

This package contains methods for fitting Generalized Linear Models (GLMs) and Generalized Additive Models (GAMs). Generalized regression models are common methods for handling data for which assuming Gaussian-distributed errors is not appropriate. For instance, if the response of interest is binary, count, or proportion data, one can instead model the expectation of the response based on an appropriate data-generating distribution. This package provides methods for fitting GLMs and GAMs under Beta regression, Poisson regression, Gamma regression, and Binomial regression (currently GLM only) settings. Models are fit using local scoring algorithms described in Hastie and Tibshirani (1990) <doi:10.1214/ss/1177013604>.

This package provides a ggplot2 geom and position for visualizing brain region data on cortical, subcortical, and white matter tract atlases. Brain atlas geometries are stored as polygon data, enabling seamless integration with the ggplot2 ecosystem including faceting, custom scales, and themes. Mowinckel & Vidal-Piñeiro (2020) <doi:10.1177/2515245920928009>.

This package provides a collection of functions to set up Google Public Data Explorer <https://www.google.com/publicdata/> data visualization tool with your own data, building automatically the corresponding DataSet Publishing Language file, or DSPL (XML), metadata file jointly with the CSV files. All zip-up and ready to be published in Public Data Explorer'.

Owing to the rich shapes of Generalised Lambda Distributions (GLDs), GLD standard/quantile/Accelerated Failure Time (AFT) regression is a competitive flexible model compared to standard/quantile/AFT regression. The proposed method has some major advantages: 1) it provides a reference line which is very robust to outliers with the attractive property of zero mean residuals and 2) it gives a unified, elegant quantile regression model from the reference line with smooth regression coefficients across different quantiles. For AFT model, it also eliminates the needs to try several different AFT models, owing to the flexible shapes of GLD. The goodness of fit of the proposed model can be assessed via QQ plots and Kolmogorov-Smirnov tests and data driven smooth test, to ensure the appropriateness of the statistical inference under consideration. Statistical distributions of coefficients of the GLD regression line are obtained using simulation, and interval estimates are obtained directly from simulated data. References include the following: Su (2015) "Flexible Parametric Quantile Regression Model" <doi:10.1007/s11222-014-9457-1>, Su (2021) "Flexible parametric accelerated failure time model"<doi:10.1080/10543406.2021.1934854>.

This General Regression Neural Networks Package uses various distance functions. It was motivated by Specht (1991, ISBN:1045-9227), and updated from previous published paper Li et al. (2016) <doi:10.1016/j.palaeo.2015.11.005>. This package includes various functions, although "euclidean" distance is used traditionally.

Quantification, analysis, and visualization of urban greenness within city networks using data from OpenStreetMap <https://www.openstreetmap.org>.

Gradient-Enhanced Kriging as an emulator for computer experiments based on Maximum-Likelihood estimation.

Several methods may be found for selecting a subset of regressors from a set of k candidate variables in multiple linear regression. One possibility is to evaluate all possible regression models and comparing them using Mallows's Cp statistic (Cp) according to Gilmour original study. Full model is calculated, all possible combinations of regressors are generated, adjusted Cp for each submodel are computed, and the submodel with the minimum adjusted value Cp (ModelMin) is calculated. To identify the final model, the package applies a sequence of hypothesis tests on submodels nested within ModelMin, following the approach outlined in Gilmour's original paper. For more details see the help of the function final_model() and the original study (1996) <doi:10.2307/2348411>.

This package provides a model building procedure to build parsimonious geoadditive model from a large number of covariates. Continuous, binary and ordered categorical responses are supported. The model building is based on component wise gradient boosting with linear effects, smoothing splines and a smooth spatial surface to model spatial autocorrelation. The resulting covariate set after gradient boosting is further reduced through backward elimination and aggregation of factor levels. The package provides a model based bootstrap method to simulate prediction intervals for point predictions. A test data set of a soil mapping case study in Berne (Switzerland) is provided. Nussbaum, M., Walthert, L., Fraefel, M., Greiner, L., and Papritz, A. (2017) <doi:10.5194/soil-3-191-2017>.

This package performs statistical data analysis of various Plant Breeding experiments. Contains functions for Line by Tester analysis as per Arunachalam, V.(1974) <http://repository.ias.ac.in/89299/> and Diallel analysis as per Griffing, B. (1956) <https://www.publish.csiro.au/bi/pdf/BI9560463>.

An R interface to weighted nonlinear least-squares optimization with the GNU Scientific Library (GSL), see M. Galassi et al. (2009, ISBN:0954612078). The available trust region methods include the Levenberg-Marquardt algorithm with and without geodesic acceleration, the Steihaug-Toint conjugate gradient algorithm for large systems and several variants of Powell's dogleg algorithm. Multi-start optimization based on quasi-random samples is implemented using a modified version of the algorithm in Hickernell and Yuan (1997, OR Transactions). Robust nonlinear regression can be performed using various robust loss functions, in which case the optimization problem is solved by iterative reweighted least squares (IRLS). Bindings are provided to tune a number of parameters affecting the low-level aspects of the trust region algorithms. The interface mimics R's nls() function and returns model objects inheriting from the same class.