Studies the reliability and generalizability of subjective judgments produced by large language models (LLMs), including annotation, rating, and structured LLM-as-a-judge tasks. Specifies evaluator, prompt, generation, and repeated-run facets through crossed or explicitly nested random sources with configurable item interactions. Fits univariate models, joint Gaussian models, and joint discrete models for binary, ordinal, and unordered categorical outcomes, with source-specific covariance. Gaussian models use exact balanced likelihood; discrete models use a dense first-order Laplace approximation with Gaussian latent random effects. Supported balanced decision studies compare evaluator, prompt, and replication allocations using observed Gaussian or explicitly requested latent binary and ordinal reliability, with random or fixed facets after Brennan (2001). Gaussian fits report asymptotic Wald standard errors for their variance components and delta-method intervals for the coefficients; discrete fits report point estimates only. Scalar nominal reliability and joint Gaussian-discrete fitting are not implemented. Includes three publicly archived LLM annotation datasets covering hate-speech, mental-health, and drug-review tasks. Discrete fitting is limited to small models; the preflight report describes supported designs and computational limits. Generalizability coefficients follow the variance-decomposition framework of Brennan (2001) &lt;<a href="https://doi.org/10.1007%2F978-1-4757-3456-0" target="_top">doi:10.1007/978-1-4757-3456-0</a>&gt;.

Extracts reproducible geometric and urban-morphology descriptors from polygon and multipart polygon building footprints stored as sf objects. Methods include area, perimeter, compactness, shape, bounding-box, topology, convexity, and orientation measures, together with optional building-height, floor-count, gross-floor-area, floor-space-index, coverage, volume, and floor-height metrics derived from explicitly named attribute columns. The resulting sf objects preserve the source attributes and geometry for spatial analysis, visualization, and downstream statistical or machine-learning workflows, with optional future-based parallel processing for larger datasets.

Implements a class of likelihood-informed frequentist dynamic borrowing methods for hybrid-control survival trials based on penalized Cox partial likelihood estimation. Implements four likelihood-informed penalty structures (precision-weighted L1, smoothed integrated-gate, information-adaptive minimax concave penalty (MCP), and likelihood-ratio-weighted L1), together with the adaptive lasso borrowing approach of Li et al. (2023, &lt;<a href="https://doi.org/10.1002%2Fbimj.202100406" target="_top">doi:10.1002/bimj.202100406</a>&gt;). Provides conditional model-based standard errors and local plug-in sandwich variance approximations, with smoothed penalties. Tools for design-stage lambda calibration via simulation, including a two-stage coarse-fine grid search, drift-level early stopping, and per-method tuning under both inference types, are also provided. A simulation harness for evaluating type I error and statistical power across population drift scenarios is included.

Read, transform, and visualize log files produced by 'Daisy' simulations. Supports importing dlf output files, reshaping depth-dependent and time-series data, and creating static and interactive plots for exploratory analysis and comparison of simulation results.

Fits species distribution models and maps habitat suitability at up to 10 m resolution from occurrence records alone, using the 'AlphaEarth' Foundations satellite embeddings (Brown et al. 2025) &lt;<a href="https://doi.org/10.48550%2FarXiv.2507.22291" target="_top">doi:10.48550/arXiv.2507.22291</a>&gt;. The embeddings, 64 values per pixel per year from a geospatial foundation model, replace environmental layers, so none need to be sourced or aligned. Provides tools to format occurrence records, place pseudo-absences, train and evaluate an ensemble of machine learning models, and export habitat-suitability rasters. Sampling, model training and prediction all run on 'Google Earth Engine', which requires a free account for noncommercial use.

This R package connects to SWI-Prolog, &lt;<a href="https://www.swi-prolog.org/" target="_top">https://www.swi-prolog.org/</a>&gt;, so that R can send deterministic and non-deterministic queries to prolog (consult, query/submit, once, findall).

Functionality for segmenting individual trees from a forest stand scanned with a close-range (e.g., terrestrial or mobile) laser scanner. The complete workflow from a raw point cloud to a complete tabular forest inventory is provided. The package contains several algorithms for detecting tree bases and a graph-based algorithm to attach all remaining points to these tree bases. It builds heavily on the 'lidR' package. A description of the segmentation algorithm can be found in Larysch et al. (2025) &lt;<a href="https://doi.org/10.1007%2Fs10342-025-01796-z" target="_top">doi:10.1007/s10342-025-01796-z</a>&gt;.

Converts between JSON text and ordinary R vectors and lists through a small, predictable set of functions, backed by vendored 'yyjson' &lt;<a href="https://github.com/ibireme/yyjson" target="_top">https://github.com/ibireme/yyjson</a>&gt; and requiring no system JSON library. Parsing accepts character, raw and file input and reports failures through structured conditions; serialization writes UTF-8 bytes suitable for use directly as an HTTP request body. The type mapping is deliberately narrow and fully documented, so what goes in and what comes out are both predictable.

Provides tools for rendering and visualizing Braille patterns in R, using the liblouis translation library. For more information see &lt;<a href="https://github.com/wininger/yaleBraille" target="_top">https://github.com/wininger/yaleBraille</a>&gt;.

Historical weather observations and food-establishment conviction records from Quebec for teaching data import, missing values, exploratory analysis and reproducible reporting. Includes bilingual prompts and feedback for a penguin data-cleaning exercise, and an optional static R code-style diagnostic based on 'lintr', as described by Hester and others (2025) &lt;<a href="https://doi.org/10.21105%2Fjoss.07240" target="_top">doi:10.21105/joss.07240</a>&gt;. Data are distributed as fixed teaching snapshots and require no network access during use.

Fits and compares representations of time-varying data against a prediction target. Given a table of targets and a table of time-stamped series belonging to them, it builds each candidate representation, from the record unreduced through a calendar grain such as a week or a month to a lookback anchored on each target, fits the requested learners on each, scores every candidate on one set of held-out folds, and stacks the out-of-fold predictions into an ensemble. Calendar-aware binning keeps a bin a real week or month rather than a fixed block of hours. Learners, response heads and metrics are registered rather than hard-coded, so adding one is a registration and not a fork of the fitting code. The penalised baseline is an elastic net fitted by cyclic coordinate descent along a warm-started path, following Friedman, Hastie and Tibshirani (2010) &lt;<a href="https://doi.org/10.18637%2Fjss.v033.i01" target="_top">doi:10.18637/jss.v033.i01</a>&gt;. The shipped default is presence-absence with a joint multi-label head scored by the true skill statistic of Allouche, Tsoar and Kadmon (2006) &lt;<a href="https://doi.org/10.1111%2Fj.1365-2664.2006.01214.x" target="_top">doi:10.1111/j.1365-2664.2006.01214.x</a>&gt;, the setting used for species distribution modelling from microclimate loggers.

Download table metadata and data from the Romanian National Institute of Statistics' TEMPO Online database. Tables can be listed in Romanian or English and downloaded as CSV files.

Hierarchical slicer-style input for Shiny applications with multi-selection, tri-state checkboxes, search, cascading selection, and a structured selection value for analytics workflows.

Provides tools for importing, processing, visualizing, fitting, comparing, and interpreting in vitro rumen gas production data. Supports ANKOM RF workflows, generic gas production datasets, and pressure-based measurements. Includes multiple kinetic models, custom nonlinear models, model comparison workflows, treatment-level ranking, diagnostic tools, and visualization functions for rumen fermentation studies.

Provides functions for designing multiple dependent state sampling inspection plans for time-truncated life tests. The package determines the minimum sample size required to satisfy a specified consumer's risk constraint and evaluates the probability of acceptance under different quality and termination ratios. Users can directly provide failure probabilities, allowing the sampling plan to be applied to different lifetime distributions without requiring distribution-specific functions. Provide operating characteristic analysis, sample size analysis, and graphical comparison with single sampling inspection plans. Aslam et al. (2016) &lt;<a href="https://doi.org/10.1080%2F08982112.2015.1068331" target="_top">doi:10.1080/08982112.2015.1068331</a>&gt;; Rao et al. (2020) &lt;<a href="https://doi.org/10.1080%2F25742558.2020.1857915" target="_top">doi:10.1080/25742558.2020.1857915</a>&gt;; Balamurali et al. (2017) &lt;<a href="https://doi.org/10.1080%2F07474946.2016.1275459" target="_top">doi:10.1080/07474946.2016.1275459</a>&gt;; Saha et al. (2021) &lt;<a href="https://doi.org/10.1080%2F21681015.2021.1893843" target="_top">doi:10.1080/21681015.2021.1893843</a>&gt;; Tripathi et al. (2020) &lt;<a href="https://doi.org/10.1007%2Fs40745-020-00267-z" target="_top">doi:10.1007/s40745-020-00267-z</a>&gt;; Tripathi et al. (2023) &lt;<a href="https://doi.org/10.1007%2Fs41872-023-00221-x" target="_top">doi:10.1007/s41872-023-00221-x</a>&gt;.

Provides tools for converting statistical results into consistent, publication-ready character output. Current functionality includes formatting p-values, confidence intervals, and correlation coefficients using publication-friendly conventions. The package emphasizes consistent reporting while preserving the underlying statistical results and statistical decision-making.

Implements principal-stratification methods for estimating time-specific and pooled heterogeneous treatment effects in longitudinal studies where outcomes may be truncated by death. Supports continuous and binary outcomes, explicit data validation and standardization, and covariate-dependent treatment effects. Fits propensity-score, principal-score, and outcome models and provides subject-level bootstrap inference, covariate-balance diagnostics, principal-stratum summaries, treatment-group-specific survival odds ratios, and outcome-noise sensitivity analysis. Methodological background is provided in &lt;<a href="https://doi.org/10.48550%2FarXiv.2608.06654" target="_top">doi:10.48550/arXiv.2608.06654</a>&gt;.

Provides goodness-of-fit tests for the skew-normal distribution with estimated parameters. Implements Kolmogorov-Smirnov and Cramér-von Mises tests using parametric bootstrap or precomputed simulation quantiles, together with robust parameter estimation procedures. Package methods and documentation are described by Li and Khang (2026) &lt;<a href="https://github.com/Divo-Lee/PBGoF" target="_top">https://github.com/Divo-Lee/PBGoF</a>&gt;.

Clean, prepare, and aggregate surveillance data for public health analysis. Provides structural data cleaning and standardisation (clean_the_nest()), age categorisation against ~50 published schemes with publication-ready labelling (preening()), time-unit aggregation with zero-filling and seasonal awareness (roost()), joint aggregation of several linked event dates (e.g. onset, admission, ICU, complication, fatality) into one table of comparable rate columns (flyway()), under-ascertainment correction via a stratified, time-varying multiplier factor supplied directly, derived by the ratio (multiplier) method, or derived by inverting an externally sourced severity rate (e.g. an infection-fatality-rate anchor) against an observed severity ratio (corncrake()), comorbidity detection from ICD-10-AM clinical coding (plumage()), vaccine coverage data construction (brood()), hash-based de-identification (molting()), and relinking of previously de-identified data (homing()). brood() produces a brood_df object supporting two population models: pre-aggregated denominators (population_model = "pre_aggregated") and record-level cohort designs (population_model = "cohort"). The cohort model handles single time-point coverage snapshots, interrupted time series analysis via a built-in sweep returning monthly coverage rates (time_series = TRUE), and birth cohort designs with person-time computation. This cohort/time-series coverage model was applied in Roughan et al. (2026) &lt;<a href="https://doi.org/10.33321%2Fcdi.2026.50.031" target="_top">doi:10.33321/cdi.2026.50.031</a>&gt; to estimate infant immunisation coverage against respiratory syncytial virus over an 18-month period. Both wide format (one row per person with dose columns, from 'starling'::murmuration()) and long format (one row per dose) are accepted. corncrake() returns both a point-corrected count and uncertainty bounds wherever they can be derived, including the inverse relationship between a severity-anchored factor and the bounds of its own reference rate. Built for Australian public health surveillance practice but not specific to it &ndash; see individual function documentation for notes on non-Australian use (e.g. Northern Hemisphere season boundaries).

Download and work with the Liking Rating Database, a curated collection of subjective liking ratings from published decision-making studies, described in Fernandez, Goyal and Krajbich (2026) &lt;<a href="https://doi.org/10.5281%2Fzenodo.22216442" target="_top">doi:10.5281/zenodo.22216442</a>&gt;. Data is read from versioned release files and cached for the session, so a pinned version returns the same rows regardless of when it is run. Provides access by dataset, by item across studies, or as the whole corpus, together with the metadata and citations needed to report it.

Estimate and forecast Bayesian simultaneous equation models for macroeconomic time series. Provides tools to specify systems of behavioral equations and accounting identities, transform and manage time series, simulate from the posterior using a Metropolis-within-Gibbs sampler, and generate unconditional and conditional forecasts with user-defined priors and restrictions. Methods are described in Rathke A. and Sarferaz S. (forthcoming) "Bayesian Estimation of Simultaneous Equations Model".

A 'shiny' application for anonymising files exported from the 'Familias' software for forensic kinship analysis (Egeland et al. (2000) &lt;<a href="https://doi.org/10.1016%2Fs0379-0738%2800%2900147-x" target="_top">doi:10.1016/s0379-0738(00)00147-x</a>&gt;). Pedigrees, marker data, allele frequencies and mutation models can be masked or modified, with options for preserving likelihood ratios exactly. The application is built on the 'pedsuite' packages for pedigree analysis.

Curates a comprehensive and robust collection of 15 classic educational, economic, and social science datasets meticulously tailored for empirical research, project-based learning, and academic instruction in higher education. It streamlines exploratory data analysis, business intelligence modeling, regression techniques, and hypothesis testing by providing ready-to-use data structures sourced from prominent community packages (Kleiber and Zeileis (2008) &lt;<a href="https://CRAN.R-project.org/package=AER" target="_top">https://CRAN.R-project.org/package=AER</a>&gt;; Fox and Weisberg (2019) &lt;<a href="https://CRAN.R-project.org/package=carData" target="_top">https://CRAN.R-project.org/package=carData</a>&gt;).

Fast, flexible framework for implementing Bayesian optimization of model hyperparameters according to the methods described in Snoek et al. (2012) &lt;<a href="https://doi.org/10.48550%2FarXiv.1206.2944" target="_top">doi:10.48550/arXiv.1206.2944</a>&gt;. The package allows the user to run scoring function in parallel, save intermediary results, and tweak other aspects of the process to fully utilize the computing resources available to the user.

Provides a comprehensive Shiny-based graphical user interface for conducting a wide range of factor analysis procedures. 'FAfA' (Factor Analysis for All) guides users through data uploading, assumption checking (descriptives, collinearity, multivariate normality, outliers), data wrangling (variable exclusion, data splitting), factor retention analysis (e.g., Parallel Analysis, Hull method, EGA), Exploratory Factor Analysis (EFA) with various rotation and extraction methods, internal split-sample EFA replication analysis, Confirmatory Factor Analysis (CFA) for model testing, Reliability Analysis (e.g., Cronbach's Alpha, McDonald's Omega), Measurement Invariance testing across groups, and item weighting techniques. The application leverages established R packages such as 'lavaan' and 'psych' to perform these analyses, offering an accessible platform for researchers and students. Results are presented in user-friendly tables and plots, with options for downloading outputs. Analysis projects can be saved and restored, and reproducible R, HTML, and PDF workflow reports can be generated.