Assessment of the distributions of baseline continuous and categorical variables in randomised trials. This method is based on the Carlisle-Stouffer method with Monte Carlo simulations. It calculates p-values for each trial baseline variable, as well as combined p-values for each trial - these p-values measure how compatible are distributions of trials baseline variables with random sampling. This package also allows for graphically plotting the cumulative frequencies of computed p-values. Please note that code was partly adapted from Carlisle JB, Loadsman JA. (2017) <doi:10.1111/anae.13650>.

Processor for selected ion flow tube mass spectrometer (SIFT-MS) output file from breath analysis. It allows the filtering of the SIFT output file (i.e., variation over time of the target analyte concentration) and the following analysis for the determination of: maximum, average, and standard deviation value of target concentration measured at each exhalation, and the respiratory rate over the measurement. Additionally, it is possible to align the SIFT-MS data with other on-line techniques such as cardio pulmonary exercise test (CPET) for a comprehensive characterization of breath samples.

Identification of Latent Patient Phenotype from Electronic Health Records (EHR) Data using Variational Bayes Gaussian Mixture Model for Latent Class Analysis and Variational Bayes regression for Biomarker level shifts, both implemented by Coordinate Ascent Variational Inference algorithms. Variational methods are used to enable Bayesian analysis of very large Electronic Health Records data. For VB GMM details see Bishop (2006,ISBN:9780-387-31073-2). For Logistic VB see Jaakkola and Jordan (2000) <doi:10.1023/A:1008932416310>. Please see preprint of JSS-submitted paper <doi:10.48550/arXiv.2512.14272>.

This package provides a flexible method for fitting regression models that can be used to find genes that are differentially expressed along one or multiple lineages in a trajectory. Based on the fitted models, it uses a variety of tests suited to answer different questions of interest, e.g. the discovery of genes for which expression is associated with pseudotime, or which are differentially expressed (in a specific region) along the trajectory. It fits a negative binomial generalized additive model (GAM) for each gene, and performs inference on the parameters of the GAM.

Dropout events make the lowly expressed genes indistinguishable from true zero expression and different than the low expression present in cells of the same type. This issue makes any subsequent downstream analysis difficult. ccImpute is an imputation algorithm that uses cell similarity established by consensus clustering to impute the most probable dropout events in the scRNA-seq datasets. ccImpute demonstrated performance which exceeds the performance of existing imputation approaches while introducing the least amount of new noise as measured by clustering performance characteristics on datasets with known cell identities.

This package implements Goldberg's (2006) <doi:10.1016/j.jrp.2006.01.001> bass-ackwards method and modern descendants for hierarchical structural analysis. Extracts solutions from 1 to k factors using principal component analysis (PCA), exploratory factor analysis (EFA), or exploratory structural equation modeling (ESEM) engines, then characterizes the hierarchy via between-level factor-score correlations computed via exact linear algebra (Waller, 2007, <doi:10.1016/j.jrp.2006.08.005>) or materialized scores. Includes the Forbes (2023) <doi:10.1037/met0000546> extension for redundancy pruning and all-levels cross-correlations.

Fits Lasso paths for high-dimensional regression using coordinate descent with automatic, data-driven tuning of the regularization parameter. The implementation is 10 to 200 times faster than the standard glmnet implementation of Lasso tuned via Cross Validation and over 100 times faster than scaled Lasso. It also provides a reliable estimate of the regression noise level and a new diagnostic for sparsity. For details of the method, see Sadhukhan, Wilms, Smeekes and Basu (2025) "Autotune: fast, accurate, and automatic tuning parameter selection for Lasso" <doi:10.48550/arXiv.2512.11139>.

This package provides a simple interface to switch between two methods for calculating basket trial characteristics, numerical integration ("exact") and Monte Carlo simulation ("simulated") for the basket trial design by Fujikawa et al. 2020 <doi:10.1002/bimj.201800404>. The exact implementation is from the baskexact package, see Baumann (2024) <doi:10.1016/j.softx.2024.101793>. The simulated implementation is from the basksim package, which was developed for Baumann et al. (2024) <doi:10.1080/19466315.2024.2402275>. The package's syntax is compatible with the basksim syntax and easily extendable.

This package provides a user friendly, easy to understand way of doing event history regression for marginal estimands of interest, including the cumulative incidence and the restricted mean survival, using the pseudo observation framework for estimation. For a review of the methodology, see Andersen and Pohar Perme (2010) <doi:10.1177/0962280209105020> or Sachs and Gabriel (2022) <doi:10.18637/jss.v102.i09>. The interface uses the well known formulation of a generalized linear model and allows for features including plotting of residuals, the use of sampling weights, and corrected variance estimation.

Fast and easy computation of Euclidean Minimum Spanning Trees (EMST) from data, relying on the R API for mlpack - the C++ Machine Learning Library (Curtin et. al., 2013). emstreeR uses the Dual-Tree Boruvka (March, Ram, Gray, 2010, <doi:10.1145/1835804.1835882>), which is theoretically and empirically the fastest algorithm for computing an EMST. This package also provides functions and an S3 method for readily visualizing Minimum Spanning Trees (MST) using either the style of the base', scatterplot3d', or ggplot2 libraries; and functions to export the MST output to shapefiles.

This package implements a Fellegi-Sunter probabilistic record linkage model that allows for missing data and the inclusion of auxiliary information. This includes functionalities to conduct a merge of two datasets under the Fellegi-Sunter model using the Expectation-Maximization algorithm. In addition, tools for preparing, adjusting, and summarizing data merges are included. The package implements methods described in Enamorado, Fifield, and Imai (2019) Using a Probabilistic Model to Assist Merging of Large-scale Administrative Records <doi:10.1017/S0003055418000783> and is available at <https://imai.fas.harvard.edu/research/linkage.html>.

Work with Microsoft Foundry from data-frame-oriented R workflows. Provides data-frame-returning helpers for Azure AI Content Safety', Azure OpenAI Responses API calls, strict structured extraction, vector representations, files, batch jobs, audio, media, and chat completions. Supports research annotation, safety gates, semantic search, and tidymodels recipes. Helps teams keep model workflows inside their Azure environment while preserving analyzable outputs. See the Microsoft Foundry REST API documentation <https://learn.microsoft.com/rest/api/microsoft-foundry/> and Azure AI Content Safety documentation <https://learn.microsoft.com/azure/ai-services/content-safety/>.

This package provides Bayesian linear and generalized linear model fitting with independent and identically distributed (iid) posterior samples. The main functions mirror R's lm() and glm() interfaces while adding prior family specifications for Gaussian, Poisson, binomial, and Gamma models with log-concave likelihoods. Sampling for supported non-conjugate models uses accept-reject methods based on likelihood subgradients as in Nygren and Nygren (2006) <doi:10.1198/016214506000000357>. The package also includes tools for prior setup, posterior summaries, prediction, diagnostics, simulation, vignettes, and optional OpenCL acceleration for larger models.

An interval-valued extension of ordinary and simple kriging. Optimization of the function is based on a generalized interval distance. This creates a non-differentiable cost function that requires a differentiable approximation to the absolute value function. This differentiable approximation is optimized using a Newton-Raphson algorithm with a penalty function to impose the constraints. Analyses in the package are driven by the intsp and intgrd classes, which are interval-valued extensions of SpatialPointsDataFrame and SpatialPixelsDataFrame respectively. The package includes several wrappers to functions in the gstat and sp packages.

This package provides functions to standardize and whiten data, and to perform Principal Component Analysis (PCA). The main advantage of this package over alternatives like prcomp() is, that jvcoords makes it easy to convert (additional) data between the original and the transformed coordinates. The package also provides a class coords, which can represent affine coordinate transformations. This class forms the basis of the transformations provided by the package, but can also be used independently. The implementation has been optimized to be of comparable speed (and sometimes even faster) than existing alternatives.

Implementation of a theoretically supported alternative to k-nearest neighbors for functional data to solve problems of estimating unobserved segments of a partially observed functional data sample, functional classification and outlier detection. The approximating neighbor curves are piecewise functions built from a functional sample. Instead of a distance on a function space we use a locally defined distance function that satisfies stabilization criteria. The package allows the implementation of the methodology and the replication of the results in Elà as, A., Jiménez, R. and Yukich, J. (2020) <arXiv:2007.16059>.

Calculate MyoScore, a genetically informed muscle health score, from bulk RNA sequencing (RNA-seq) raw count data. MyoScore integrates results from genome-wide association studies (GWAS) and transcriptome-wide association studies (TWAS) across 28 muscle-related phenotypes to quantify muscle health along five dimensions (Strength, Mass, LeanMuscle, Youth, Resilience), each scored from 0 to 100. The package provides preprocessing via counts per million (CPM) normalization, dimension-level and composite scoring, and visualization utilities including radar charts and grouped boxplots. For more information, see <https://github.com/Hirriririir/MyoScore>.

Comprehensive analytical tools are provided to characterize infectious disease superspreading from contact tracing surveillance data. The underlying theoretical frameworks of this toolkit include branching process with transmission heterogeneity (Lloyd-Smith et al. (2005) <doi:10.1038/nature04153>), case cluster size distribution (Nishiura et al. (2012) <doi:10.1016/j.jtbi.2011.10.039>, Blumberg et al. (2014) <doi:10.1371/journal.ppat.1004452>, and Kucharski and Althaus (2015) <doi:10.2807/1560-7917.ES2015.20.25.21167>), and decomposition of reproduction number (Zhao et al. (2022) <doi:10.1371/journal.pcbi.1010281>).

Pupillometric data collected using SR Research Eyelink eye trackers requires significant preprocessing. This package contains functions for preparing pupil dilation data for visualization and statistical analysis. Specifically, it provides a pipeline of functions which aid in data validation, the removal of blinks/artifacts, downsampling, and baselining, among others. Additionally, plotting functions for creating grand average and conditional average plots are provided. See the vignette for samples of the functionality. The package is designed for handling data collected with SR Research Eyelink eye trackers using Sample Reports created in SR Research Data Viewer.

This package provides functions for conducting power analysis in ANOVA designs, including between-, within-, and mixed-factor designs, with full support for both main effects and interactions. The package allows calculation of statistical power, required total sample size, significance level, and minimal detectable effect sizes expressed as partial eta squared or Cohen's f for ANOVA terms and planned contrasts. In addition, complementary functions are included for common related tests such as t-tests and correlation tests, making the package a convenient toolkit for power analysis in experimental psychology and related fields.

Traditional model evaluation metrics fail to capture model performance under less than ideal conditions. This package employs techniques to evaluate models "under-stress". This includes testing models extrapolation ability, or testing accuracy on specific sub-samples of the overall model space. Details describing stress-testing methods in this package are provided in Haycock (2023) <doi:10.26076/2am5-9f67>. The other primary contribution of this package is provided to R users access to the Python library PyCaret <https://pycaret.org/> for quick and easy access to auto-tuned machine learning models.

The statistical tools in this package do one of four things: 1) Enhance basic statistical functions with more flexible inputs, smarter defaults, and richer, clearer, and ready-to-use output (e.g., t.test2()) 2) Produce publication-ready commonly needed figures with one line of code (e.g., plot_cdf()) 3) Implement novel analytical tools developed by the authors (e.g., twolines()) 4) Deliver niche functions of high value to the authors that are not easily available elsewhere (e.g., clear(), convert_to_sql(), resize_images()).

This package provides a latent, quasi-independent truncation time is assumed to be linked with the observed dependent truncation time, the event time, and an unknown transformation parameter via a structural transformation model. The transformation parameter is chosen to minimize the conditional Kendall's tau (Martin and Betensky, 2005) <doi:10.1198/016214504000001538> or the regression coefficient estimates (Jones and Crowley, 1992) <doi:10.2307/2336782>. The marginal distribution for the truncation time and the event time are completely left unspecified. The methodology is applied to survival curve estimation and regression analysis.

Create adjacency matrices of vocalisation graphs from dataframes containing sequences of speech and silence intervals, transforming these matrices into Markov diagrams, and generating datasets for classification of these diagrams by flattening them and adding global properties (functionals) etc. Vocalisation diagrams date back to early work in psychiatry (Jaffe and Feldstein, 1970) and social psychology (Dabbs and Ruback, 1987) but have only recently been employed as a data representation method for machine learning tasks including meeting segmentation (Luz, 2012) <doi:10.1145/2328967.2328970> and classification (Luz, 2013) <doi:10.1145/2522848.2533788>.