Extension of funHDDC Schmutz et al. (2018) <doi:10.1007/s00180-020-00958-4> for cases including outliers by fitting t-distributions for robust groups. TFunHDDC can cluster univariate or multivariate data produced by the fda package for data using a b-splines or Fourier basis.

This package provides functions for multivariate analysis with compositional data. Includes a function for doing compositional canonical correlation analysis. This analysis requires two data matrices of compositions, which can be adequately transformed and used as entries in a specialized program for canonical correlation analysis, that is able to deal with singular covariance matrices. The methodology is described in Graffelman et al. (2017) <doi:10.1101/144584>. Functions for log-ratio principal component analysis with condition number computations and log-ratio discriminant analysis have been added to the package.

Hospitals, hospital systems, and even trauma systems that provide care to injured patients may not be aware of robust metrics that can help gauge the efficacy of their programs in saving the lives of injured patients. traumar provides robust functions driven by the academic literature to automate the calculation of relevant metrics to individuals desiring to measure the performance of their trauma center or even a trauma system. traumar also provides some helper functions for the data analysis journey. Users can refer to the following publications for descriptions of the methods used in traumar'. TRISS methodology, including probability of survival, and the W, M, and Z Scores - Flora (1978) <doi:10.1097/00005373-197810000-00003>, Boyd et al. (1987, PMID:3106646), Llullaku et al. (2009) <doi:10.1186/1749-7922-4-2>, Singh et al. (2011) <doi:10.4103/0974-2700.86626>, Baker et al. (1974, PMID:4814394), and Champion et al. (1989) <doi:10.1097/00005373-198905000-00017>. For the Relative Mortality Metric, see Napoli et al. (2017) <doi:10.1080/24725579.2017.1325948>, Schroeder et al. (2019) <doi:10.1080/10903127.2018.1489021>, and Kassar et al. (2016) <doi:10.1177/00031348221093563>. For more information about methods to calculate over- and under-triage in trauma hospital populations and samples, please see the following publications - Peng & Xiang (2016) <doi:10.1016/j.ajem.2016.08.061>, Beam et al. (2022) <doi:10.23937/2474-3674/1510136>, Roden-Foreman et al. (2017) <doi:10.1097/JTN.0000000000000283>.

Converting text to numerical features requires specifically created procedures, which are implemented as steps according to the recipes package. These steps allows for tokenization, filtering, counting (tf and tfidf) and feature hashing.

Autoregressive distributed lag (A[R]DL) models (and their reparameterized equivalent, the Generalized Error-Correction Model [GECM]) are the workhorse models in uncovering dynamic inferences. ADL models are simple to estimate; this is what makes them attractive. Once these models are estimated, what is less clear is how to uncover a rich set of dynamic inferences from these models. We provide tools for recovering those inferences. These tools apply to traditional time-series quantities of interest and are built from the Impulse Response Function and Step Response Function (sometimes described as a pulse effect or a cumulative effect). They also allow for a variety of shock histories to be applied to the independent variable (beyond just a one-time, one-unit increase) as well as the recovery of inferences in levels for shocks applied to (in)dependent variables in differences (what we call the Generalized Dynamic Response Function). These effects are also available for the general conditional dynamic model advocated by Warner, Vande Kamp, and Jordan (2026 <doi:10.1017/psrm.2026.10087>). We also provide the formulae for these effects.

Greedy optimal subset selection for transformation models (Hothorn et al., 2018, <doi:10.1111/sjos.12291> ) based on the abess algorithm (Zhu et al., 2020, <doi:10.1073/pnas.2014241117> ). Applicable to models from packages tram and cotram'. Application to shift-scale transformation models are described in Siegfried et al. (2024, <doi:10.1080/00031305.2023.2203177>).

This package provides a standardized user interface for column selection, that facilitates dataset merging in teal framework.

Various methods for targeted and semiparametric inference including augmented inverse probability weighted (AIPW) estimators for missing data and causal inference (Bang and Robins (2005) <doi:10.1111/j.1541-0420.2005.00377.x>), one-step imputation (Nordland et al (2026)) <doi:10.48550/arXiv.2606.07174>), variable importance and conditional average treatment effects (CATE) (van der Laan (2006) <doi:10.2202/1557-4679.1008>), estimators for risk differences and relative risks (Richardson et al. (2017) <doi:10.1080/01621459.2016.1192546>), assumption lean inference for generalized linear model parameters (Vansteelandt et al. (2022) <doi:10.1111/rssb.12504>).

Analyze telemetry datasets generalized to allow any technology. The filtering steps check for false positives caused by reflected transmissions from surfaces and false pings from other noise generating equipment. The filters are based on JSATS filtering algorithms found in package filteRjsats <https://CRAN.R-project.org/package=filteRjsats> but have been generalized to allow the user to define many of the filtering variables. Additionally, this package contains scripts used to help identify an optimal maximum blanking period as defined in Capello et al (2015) <doi:10.1371/journal.pone.0134002>. The functions were written according to their manuscript description, but have not been reviewed by the authors for accuracy. It is included here as is, without warranty.

Easily construct prompts and associated logic for interacting with large language models (LLMs). tidyprompt introduces the concept of prompt wraps, which are building blocks that you can use to quickly turn a simple prompt into a complex one. Prompt wraps do not just modify the prompt text, but also add extraction and validation functions that will be applied to the response of the LLM. This ensures that the user gets the desired output. tidyprompt can add various features to prompts and their evaluation by LLMs, such as structured output, automatic feedback, retries, reasoning modes, autonomous R function calling, and R code generation and evaluation. It is designed to be compatible with any LLM provider that offers chat completion.

Convert T Cell Receptor (TCR) gene names between the 10X Genomics, Adaptive Biotechnologies, and ImMunoGeneTics (IMGT) nomenclatures.

This package provides a unified estimation procedure for the analysis of right censored data using linear transformation models. An introduction can be found in Jie Zhou et al. (2022) <doi:10.18637/jss.v101.i09>.

Several functions to allow comparisons of data across different geographies, in particular for Canadian census data from different censuses.

This package provides a framework to work with decision rules. Rules can be extracted from supported models, augmented with (custom) metrics using validation data, manipulated using standard dataframe operations, reordered and pruned based on a metric, predict on unseen (test) data. Utilities include; Creating a rulelist manually, Exporting a rulelist as a SQL case statement and so on. The package offers two classes; rulelist and ruleset based on dataframe.

Provide the core functionality to transform longitudinal data to complex-time (kime) data using analytic and numerical techniques, visualize the original time-series and reconstructed kime-surfaces, perform model based (e.g., tensor-linear regression) and model-free classification and clustering methods in the book Dinov, ID and Velev, MV. (2021) "Data Science: Time Complexity, Inferential Uncertainty, and Spacekime Analytics", De Gruyter STEM Series, ISBN 978-3-11-069780-3. <https://www.degruyter.com/view/title/576646>. The package includes 18 core functions which can be separated into three groups. 1) draw longitudinal data, such as Functional magnetic resonance imaging(fMRI) time-series, and forecast or transform the time-series data. 2) simulate real-valued time-series data, e.g., fMRI time-courses, detect the activated areas, report the corresponding p-values, and visualize the p-values in the 3D brain space. 3) Laplace transform and kimesurface reconstructions of the fMRI data.

Computes the t* statistic corresponding to the tau* population coefficient introduced by Bergsma and Dassios (2014) <DOI:10.3150/13-BEJ514> and does so in O(n^2) time following the algorithm of Heller and Heller (2016) <DOI:10.48550/arXiv.1605.08732> building off of the work of Weihs, Drton, and Leung (2016) <DOI:10.1007/s00180-015-0639-x>. Also allows for independence testing using the asymptotic distribution of t* as described by Nandy, Weihs, and Drton (2016) <DOI:10.1214/16-EJS1166>.

This package provides a minimal-dependency R client for OpenAI'-compatible text-to-speech APIs (see <https://developers.openai.com/api/reference/resources/audio>). Supports OpenAI', ElevenLabs', local servers ('Chatterbox', LM Studio', OpenWebUI', AnythingLLM'), and the chatterbox package for in-process synthesis. Provides a unified interface for speech synthesis, voice cloning, and voice design with backend-specific parameters.

This package implements a task queue system for asynchronous parallel computing using PostgreSQL <https://www.postgresql.org/> as a backend. Designed for embarrassingly parallel problems where tasks do not communicate with each other. Dynamically distributes tasks to workers, handles uneven load balancing, and allows new workers to join at any time. Particularly useful for running large numbers of independent tasks on high-performance computing (HPC) clusters with SLURM <https://slurm.schedmd.com/> job schedulers.

The satisfaction Analysis using the tetraclasse model from Sylvie Llosa. Llosa (1997) <http://www.jstor.org/stable/40592578>.

Split a dataframe, tibble, or data.table into training and test sets. Return either a list, an index, or directly assign training and test sets into memory.

This package provides a flexible simulation tool for phylogenetic trees under a general model for speciation and extinction. Trees with a user-specified number of extant tips, or a user-specified stem age are simulated. It is possible to assume any probability distribution for the waiting time until speciation and extinction. Furthermore, the waiting times to speciation / extinction may be scaled in different parts of the tree, meaning we can simulate trees with clade-dependent diversification processes. At a speciation event, one species splits into two. We allow for two different modes at these splits: (i) symmetric, where for every speciation event new waiting times until speciation and extinction are drawn for both daughter lineages; and (ii) asymmetric, where a speciation event results in one species with new waiting times, and another that carries the extinction time and age of its ancestor. The symmetric mode can be seen as an vicariant or allopatric process where divided populations suffer equal evolutionary forces while the asymmetric mode could be seen as a peripatric speciation where a mother lineage continues to exist. Reference: O. Hagen and T. Stadler (2017). TreeSimGM: Simulating phylogenetic trees under general Bellman Harris models with lineage-specific shifts of speciation and extinction in R. Methods in Ecology and Evolution. <doi:10.1111/2041-210X.12917>.

This package provides a new measure of similarity between a pair of mass spectrometry (MS) experiments, called truncated rank correlation (TRC). To provide a robust metric of similarity in noisy high-dimensional data, TRC uses truncated top ranks (or top m-ranks) for calculating correlation. Truncated rank correlation as a robust measure of test-retest reliability in mass spectrometry data. For more details see Lim et al. (2019) <doi:10.1515/sagmb-2018-0056>.

The strength of evidence provided by epidemiological and observational studies is inherently limited by the potential for unmeasured confounding. We focus on three key quantities: the observed bound of the confidence interval closest to the null, the relationship between an unmeasured confounder and the outcome, for example a plausible residual effect size for an unmeasured continuous or binary confounder, and the relationship between an unmeasured confounder and the exposure, for example a realistic mean difference or prevalence difference for this hypothetical confounder between exposure groups. Building on the methods put forth by Cornfield et al. (1959), Bross (1966), Schlesselman (1978), Rosenbaum & Rubin (1983), Lin et al. (1998), Lash et al. (2009), Rosenbaum (1986), Cinelli & Hazlett (2020), VanderWeele & Ding (2017), and Ding & VanderWeele (2016), we can use these quantities to assess how an unmeasured confounder may tip our result to insignificance.

Transfer learning for generalized factor models with support for continuous, count (Poisson), and binary data types. The package provides functions for single and multiple source transfer learning, source detection to identify positive and negative transfer sources, factor decomposition using Maximum Likelihood Estimation (MLE), and information criteria ('IC1 and IC2') for rank selection. The methods are particularly useful for high-dimensional data analysis where auxiliary information from related source datasets can improve estimation efficiency in the target domain.