This package provides a collection of fast statistical and utility functions for data analysis. Functions for regression, maximum likelihood, column-wise statistics and many more have been included. C++ has been utilized to speed up the functions. References: Tsagris M., Papadakis M. (2018). Taking R to its limits: 70+ tips. PeerJ Preprints 6:e26605v1 <doi:10.7287/peerj.preprints.26605v1>.

Flexible statistical modelling using a modular framework for regression, in which groups of transformations are composed together and act on probability distributions.

After defining an R6 class, R62S3 is used to automatically generate optional S3/S4 generics and methods for dispatch. Also allows piping for R6 objects.

Fit statistical models based on the Dawid-Skene model - Dawid and Skene (1979) <doi:10.2307/2346806> - to repeated categorical rating data. Full Bayesian inference for these models is supported through the Stan modelling language. rater also allows the user to extract and plot key parameters of these models.

Developed to assist researchers with planning analysis, prior to obtaining data from Trusted Research Environments (TREs) also known as safe havens. With functionality to export and import marginal distributions as well as synthesise data, both with and without correlations from these marginal distributions. Using a multivariate cumulative distribution (COPULA). Additionally the International Stroke Trial (IST) is included as an example dataset under ODC-By licence Sandercock et al. (2011) <doi:10.7488/ds/104>, Sandercock et al. (2011) <doi:10.1186/1745-6215-12-101>.

Helps to prepare a release. Before releasing an R package it is important to update the DESCRIPTION file and the changelog. This package prepares these files and also updates the versions according to the branches. It relies heavily on the desc packages.

This package provides tools for manipulating, exploring, and visualising multiple-response data, including scored or ranked responses. Conversions to and from factors, lists, strings, matrices; reordering, lumping, flattening; set operations; tables; frequency and co-occurrence plots.

Fast tools for unequal probability sampling in multi-dimensional spaces, implemented in Rust for high performance. The package offers a wide range of methods, including Sampford (Sampford, 1967, <doi:10.1093/biomet/54.3-4.499>) and correlated Poisson sampling (Bondesson and Thorburn, 2008, <doi:10.1111/j.1467-9469.2008.00596.x>), pivotal sampling (Deville and Tillé, 1998, <doi:10.1093/biomet/91.4.893>), and balanced sampling such as the cube method (Deville and Tillé, 2004, <doi:10.1093/biomet/91.4.893>) to ensure auxiliary totals are respected. Spatially balanced approaches, including the local pivotal method (Grafström et al., 2012, <doi:10.1111/j.1541-0420.2011.01699.x>), spatially correlated Poisson sampling (Grafström, 2012, <doi:10.1016/j.jspi.2011.07.003>), and locally correlated Poisson sampling (Prentius, 2024, <doi:10.1002/env.2832>), provide efficient designs when the target variable is linked to auxiliary information.

QuantLib bindings are provided for R using Rcpp via an updated variant of the header-only Quantuccia project (put together initially by Peter Caspers) offering an essential subset of QuantLib (and now maintained separately for the calendaring subset). See the included file AUTHORS for a full list of contributors to both QuantLib and Quantuccia'. Note that this package provided an initial viability proof, current work is done (via approximately quarterly releases tracking QuantLib') in the smaller package qlcal which is generally preferred.

Providing wrapper functions to implement Bayesian analysis in JAGS. Some major features include monitoring convergence of a MCMC model using Rubin and Gelman Rhat statistics, automatically running a MCMC model till it converges, and implementing parallel processing of a MCMC model for multiple chains.

This package provides popular sampling distributions C++ routines based in armadillo through a header file approach.

An integrated package for constructing random forest prediction intervals using a fast implementation package ranger'. This package can apply the following three methods described in Haozhe Zhang, Joshua Zimmerman, Dan Nettleton, and Daniel J. Nordman (2019) <doi:10.1080/00031305.2019.1585288>: the out-of-bag prediction interval, the split conformal method, and the quantile regression forest.

This package performs the Joint and Individual Variation Explained (JIVE) decomposition on a list of data sets when the data share a dimension, returning low-rank matrices that capture the joint and individual structure of the data [O'Connell, MJ and Lock, EF (2016) <doi:10.1093/bioinformatics/btw324>]. It provides two methods of rank selection when the rank is unknown, a permutation test and a Bayesian Information Criterion (BIC) selection algorithm. Also included in the package are three plotting functions for visualizing the variance attributed to each data source: a bar plot that shows the percentages of the variability attributable to joint and individual structure, a heatmap that shows the structure of the variability, and principal component plots.

Linguistic Descriptions of Complex Phenomena (LDCP) is an architecture and methodology that allows us to model complex phenomena, interpreting input data, and generating automatic text reports customized to the user needs (see <doi:10.1016/j.ins.2016.11.002> and <doi:10.1007/s00500-016-2430-5>). The proposed package contains a set of methods that facilitates the development of LDCP systems. It main goal is increasing the visibility and practical use of this research line.

This package provides a tool for mass deployment of shiny apps to RStudio Connect or Shiny Server'. Multiple user accounts and servers can be configured for deployment.

Read and write labelled sparse matrices in text format as used by software such as SVMLight', LibSVM', ThunderSVM', LibFM', xLearn', XGBoost', LightGBM', and others. Supports labelled data for regression, classification (binary, multi-class, multi-label), and ranking (with qid field), and can handle header metadata and comments in files.

This package implements solutions to canonical models of Economics such as Monopoly Profit Maximization, Cournot's Duopoly, Solow (1956, <doi:10.2307/1884513>) growth model and Mankiw, Romer and Weil (1992, <doi:10.2307/2118477>) growth model.

Routines to interact with the Numerai Machine Learning Tournament API <https://numer.ai>. The functionality includes the ability to automatically download the current tournament data, submit predictions, and to get information for your user.

We provide a toolbox to fit and simulate a univariate or multivariate damped random walk process that is also known as an Ornstein-Uhlenbeck process or a continuous-time autoregressive model of the first order, i.e., CAR(1) or CARMA(1, 0). This process is suitable for analyzing univariate or multivariate time series data with irregularly-spaced observation times and heteroscedastic measurement errors. When it comes to the multivariate case, the number of data points (measurements/observations) available at each observation time does not need to be the same, and the length of each time series can vary. The number of time series data sets that can be modeled simultaneously is limited to ten in this version of the package. We use Kalman-filtering to evaluate the resulting likelihood function, which leads to a scalable and efficient computation in finding maximum likelihood estimates of the model parameters or in drawing their posterior samples. Please pay attention to loading the data if this package is used for astronomical data analyses; see the details in the manual. Also see Hu and Tak (2020) <arXiv:2005.08049>.

This package implements the estimation techniques described in Rousseeuw & Verboven (2002) <doi:10.1016/S0167-9473(02)00078-6> for the location and scale of very small samples.

Response surface designs with neighbour effects are suitable for experimental situations where it is expected that the treatment combination administered to one experimental unit may affect the response on neighboring units as well as the response on the unit to which it is applied (Dalal et al.,2025 <doi: 10.57805/revstat.v23i2.513>). Integrating these effects in the response surface model improves the experiment's precision Verma A., Jaggi S., Varghese, E.,Varghese, C.,Bhowmik, A., Datta, A. and Hemavathi M. (2021)<doi: 10.1080/03610918.2021.1890123>). This package includes sym(), asym1(), asym2(), asym3() and asym4() functions that generates response surface designs which are rotatable under a polynomial model of a given order without interaction term incorporating neighbour effects.

The header-only C++ template library FastAD for automatic differentiation <https://github.com/JamesYang007/FastAD> is provided by this package, along with a few illustrative examples that can all be called from R.

As of RStudio v1.3, the preferences in the Global Options dialog (and a number of other preferences that arenâ t) are now saved in simple, plain-text JSON files. This package provides an interface for working with these RStudio JSON preference files to easily make modifications without using the point-and-click option menus. This is particularly helpful when working on teams to ensure a unified experience across machines and utilizing settings for best practices.

This package performs RNA emulation and active learning proposed by Heo and Sung (2025) <doi:10.1080/00401706.2024.2376173> for multi-fidelity computer experiments. The RNA emulator is particularly useful when the simulations with different fidelity level are nonlinearly correlated. The hyperparameters in the model are estimated by maximum likelihood estimation.