This package provides Stein-discrepancy goodness-of-fit tests and Stein-method-based sampling tools. The tests include kernel Stein discrepancy U- and V-statistics following Liu et al. (2016) <doi:10.48550/arXiv.1602.03253> and Chwialkowski et al. (2016) <doi:10.48550/arXiv.1602.02964>, plus the finite set Stein discrepancy test of Jitkrittum et al. (2017) <doi:10.48550/arXiv.1705.07673>. The sampling tools include Stein thinning, Stein Points, Stein Point Markov chain Monte Carlo, and Stein variational gradient descent following Riabiz et al. (2022) <doi:10.48550/arXiv.2005.03952>, Chen et al. (2018) <doi:10.48550/arXiv.1803.10161>, Chen et al. (2019) <doi:10.48550/arXiv.1905.03673>, and Liu and Wang (2016) <doi:10.48550/arXiv.1608.04471>. Gaussian mixture utilities are included for constructing example targets, simulation, density evaluation, and score callbacks.

This package provides a user-friendly R package that enables the characterization of each cfDNA fragment overlapping one or multiple mutations of interest, starting from a sequencing file containing aligned reads (BAM file). fRagmentomics supports multiple mutation input formats (e.g., VCF, TSV, or string "chr:pos:ref:alt" representation), accommodates one-based and zero-based genomic conventions, handles mutation representation ambiguities, and accepts any reference file and species in FASTA format. For each cfDNA fragment, fRagmentomics outputs its size, its 3 and 5 sequences, and its mutational status. Optionally, when users set apply_bcftools_norm = TRUE, fRagmentomics invokes the external command-line tool bcftools norm to left-align and normalize variants. If bcftools is not found on the system PATH while this option is enabled, the function errors. The package does not install external software; see the INSTALL file for per-OS instructions.

Cancer cells accumulate DNA mutations as result of DNA damage and DNA repair processes. This computational framework is aimed at deciphering DNA mutational signatures operating in cancer. The framework includes modules that support raw data import and processing, mutational signature extraction, and results interpretation and visualization. The framework accepts widely used file formats storing information about DNA variants, such as Variant Call Format files. The framework performs Non-Negative Matrix Factorization to extract mutational signatures explaining the observed set of DNA mutations. Bootstrapping is performed as part of the analysis. The framework supports parallelization and is optimized for use on multi-core systems. The software was described by Fantini D et al (2020) <doi:10.1038/s41598-020-75062-0> and is based on a custom R-based implementation of the original MATLAB WTSI framework by Alexandrov LB et al (2013) <doi:10.1016/j.celrep.2012.12.008>.

The Genetic Algorithm (GA) is used to perform changepoint analysis in time series data. The package also includes an extended island version of GA, as described in Lu, Lund, and Lee (2010, <doi:10.1214/09-AOAS289>). By mimicking the principles of natural selection and evolution, GA provides a powerful stochastic search technique for solving combinatorial optimization problems. In changepointGA', each chromosome represents a changepoint configuration, including the number and locations of changepoints, hyperparameters, and model parameters. The package employs genetic operatorsâ selection, crossover, and mutationâ to iteratively improve solutions based on the given fitness (objective) function. Key features of changepointGA include encoding changepoint configurations in an integer format, enabling dynamic and simultaneous estimation of model hyperparameters, changepoint configurations, and associated parameters. The detailed algorithmic implementation can be found in the package vignettes and in the paper of Li and Lu (2024, <doi:10.48550/arXiv.2410.15571>).

Quantifying systematic heterogeneity in meta-analysis using R. The M statistic aggregates heterogeneity information across multiple variants to, identify systematic heterogeneity patterns and their direction of effect in meta-analysis. It's primary use is to identify outlier studies, which either show "null" effects or consistently show stronger or weaker genetic effects than average across, the panel of variants examined in a GWAS meta-analysis. In contrast to conventional heterogeneity metrics (Q-statistic, I-squared and tau-squared) which measure random heterogeneity at individual variants, M measures systematic (non-random) heterogeneity across multiple independently associated variants. Systematic heterogeneity can arise in a meta-analysis due to differences in the study characteristics of participating studies. Some of the differences may include: ancestry, allele frequencies, phenotype definition, age-of-disease onset, family-history, gender, linkage disequilibrium and quality control thresholds. See <https://magosil86.github.io/getmstatistic/> for statistical statistical theory, documentation and examples.

Pool dilution is a isotope tracer technique wherein a biogeochemical pool is artifically enriched with its heavy isotopologue and the gross productive and consumptive fluxes of that pool are quantified by the change in pool size and isotopic composition over time. This package calculates gross production and consumption rates from closed-system isotopic pool dilution time series data. Pool size concentrations and heavy isotope (e.g., 15N) content are measured over time and the model optimizes production rate (P) and the first order rate constant (k) by minimizing error in the model-predicted total pool size, as well as the isotopic signature. The model optimizes rates by weighting information against the signal:noise ratio of concentration and heavy- isotope signatures using measurement precision as well as the magnitude of change over time. The calculations used here are based on von Fischer and Hedin (2002) <doi:10.1029/2001GB001448> with some modifications.

This package performs Bayesian unit root testing for autoregressive time series models with non-linear trend components approximated by linear spline functions, as proposed by Kumar et al. (2020) <doi:10.19139/soic-2310-5070-786>. The package BayesSplineUR computes posterior odds ratios, Bayes factors, and posterior probabilities for the unit root hypothesis against trend-stationary alternatives in models with linear spline trends or maintained polynomial trends as developed by Chaturvedi and Kumar (2005) <doi:10.1016/j.spl.2005.04.044>. Includes automatic knot selection using information criteria (AIC/BIC) and theoretical foundations for Bayesian unit root testing under structural breaks and maintained trends drawing from Schotman and van Dijk (1991) <doi:10.1016/0304-4076(91)90038-F>, Phillips and Perron (1988) <doi:10.1093/biomet/75.2.335>, Ouliaris et al. (1988) <doi:10.1007/978-94-009-2953-1_10>, and Perron (1989) <doi:10.2307/1913683>.

Fit and apply ComBat, linear mixed-effects models (LMM), or prescaling to harmonize magnetic resonance imaging (MRI) data from different sites. Briefly, these methods remove differences between sites due to using different scanning devices, and LMM additionally tests linear hypotheses. As detailed in the manual, the original ComBat function was first modified for the harmonization of MRI data (Fortin et al. (2017) <doi:10.1016/j.neuroimage.2017.11.024>) and then modified again to create separate functions for fitting and applying the harmonization and allow missing values and constant rows for its use within the Enhancing Neuro Imaging Genetics through Meta-Analysis (ENIGMA) Consortium (Radua et al. (2020) <doi:10.1016/j.neuroimage.2020.116956>); this package includes the latter version. LMM calls "lme" massively considering specific brain imaging details. Finally, prescaling is a good option for fMRI, where different devices can have varying units of measurement.

This package provides functions to calculate and visualise the Regional Consistency Probability (RCP) for single-arm multi-regional clinical trials (MRCTs) using the Effect Retention Approach (ERA). Six endpoint types are supported: continuous, binary, count (negative binomial), time-to-event via hazard ratio, milestone survival, and restricted mean survival time (RMST). For each endpoint, both a closed-form (or semi-analytical) solution and a Monte Carlo simulation approach are implemented. Two consistency evaluation methods are available: Method 1 (effect retention in Region 1 relative to the overall population) and Method 2 (simultaneous positive effect across all regions). Plotting functions generate faceted visualisations of RCP as a function of the regional allocation proportion, overlaying formula and simulation results for direct comparison. The methodology follows the Japanese MHLW guidelines for MRCTs. Abbreviations used: RCP (Regional Consistency Probability), MRCT (Multi-Regional Clinical Trial), RMST (Restricted Mean Survival Time), MHLW (Ministry of Health, Labour and Welfare).

This package performs unadjusted Bayesian survival analysis for right censored time-to-event data. The main function, BayesSurv(), computes the posterior mean and a credible band for the survival function and for the cumulative hazard, as well as the posterior mean for the hazard, starting from a piecewise exponential (histogram) prior with Gamma distributed heights that are either independent, or have a Markovian dependence structure. A function, PlotBayesSurv(), is provided to easily create plots of the posterior means of the hazard, cumulative hazard and survival function, with a credible band accompanying the latter two. The priors and samplers are described in more detail in Castillo and Van der Pas (2020) "Multiscale Bayesian survival analysis" <arXiv:2005.02889>. In that paper it is also shown that the credible bands for the survival function and the cumulative hazard can be considered confidence bands (under mild conditions) and thus offer reliable uncertainty quantification.

Translates several CSV files with ontological terms and corresponding data into RDF triples. These RDF triples are stored in OWL and JSON-LD files, facilitating data accessibility, interoperability, and knowledge unification. The triples are also visualized in a graph saved as an SVG. The input CSVs must be formatted with a template from a public Google Sheet; see README or vignette for more information. This is a tool is used by the SDLE Research Center at Case Western Reserve University to create and visualize material science ontologies, and it includes example ontologies to demonstrate its capabilities. This work was supported by the U.S. Department of Energyâ s Office of Energy Efficiency and Renewable Energy (EERE) under Solar Energy Technologies Office (SETO) Agreement Numbers E-EE0009353 and DE-EE0009347, Department of Energy (National Nuclear Security Administration) under Award Number DE-NA0004104 and Contract number B647887, and U.S. National Science Foundation Award under Award Number 2133576.

This package provides a comprehensive suite of statistical tools tailored for agricultural and plant breeding research. Provides automated pipelines for analysis of variance and covariance under randomized complete block designs and completely randomized designs, descriptive summary statistics, and post-hoc multiple range tests including Least Significant Difference, Tukey, and Scheffe based on Steel et al. (1997) <isbn:978-0070610286>. Quantitative genetic parameters including genotypic, phenotypic, and environmental variance components and broad-sense heritability follow Burton and Devane (1953) <doi:10.2134/agronj1953.00021962004500100005x>. Genetic advance and genetic advance as percentage of mean estimation follow Johnson et al. (1955) <doi:10.2134/agronj1955.00021962004700070009x>. Genotypic, phenotypic, and environmental correlations follow Miller et al. (1958) <doi:10.2134/agronj1958.00021962005000100020x>. Genotypic and phenotypic path coefficient analysis direct and indirect effects decomposition follows Dewey and Lu (1959) <doi:10.2134/agronj1959.00021962005100090002x>. Principal component analysis follows Jolliffe (2002) <isbn:978-0387954424> and hierarchical clustering follows Sneath and Sokal (1973) <isbn:978-0716706977>.

Simple and user-friendly wrappers to the saemix package for performing linear and non-linear mixed-effects regression modeling for growth data to account for clustering or longitudinal analysis via repeated measurements. The package allows users to fit a variety of growth models, including linear, exponential, logistic, and Gompertz functions. For non-linear models, starting values are automatically calculated using initial least-squares estimates. The package includes functions for summarizing models, visualizing data and results, calculating doubling time and other key statistics, and generating model diagnostic plots and residual summary statistics. It also provides functions for generating publication-ready summary tables for reports. Additionally, users can fit linear and non-linear least-squares regression models if clustering is not applicable. The mixed-effects modeling methods in this package are based on Comets, Lavenu, and Lavielle (2017) <doi:10.18637/jss.v080.i03> as implemented in the saemix package. Please contact us at models@dfci.harvard.edu with any questions.

This package creates interactive 3D Burak Diagrams for evaluating climatological and hydrological simulations. The BD-Clim framework evaluates climatological simulations by representing correlation, standard deviation, centered root mean-square difference, bias, and root mean-square difference. The BD-HydNSE framework evaluates hydrological simulations using correlation, standard deviation, centered root mean-square difference, percent bias, and Nash-Sutcliffe efficiency. The BD-HydKGE framework evaluates hydrological simulations using correlation, standard deviation, centered root mean-square difference, percent bias, and Kling-Gupta efficiency. The frameworks extend the Taylor Diagram ( Taylor (2001) <doi:10.1029/2000JD900719> ) by introducing an orthogonal axis for bias and representing selected performance metrics as surfaces in the resulting 3D space. Nash-Sutcliffe efficiency was introduced by Nash and Sutcliffe (1970) <doi:10.1016/0022-1694(70)90255-6>, and Kling-Gupta efficiency was proposed by Gupta et al. (2009) <doi:10.1016/j.jhydrol.2009.08.003>. The diagrams are interactive and can optionally be saved as HTML files.

This package provides a point-and-click Shiny interface to the biocharkit package. Lets a user upload Excel workbooks of biochar characterisation and batch adsorption data, map spreadsheet columns to the required variables via dropdown menus, and run sample-ID parsing, adsorption capacity and removal efficiency calculations, isotherm fitting (Langmuir, Freundlich, Temkin, Dubinin-Radushkevich, Sips), kinetics fitting (pseudo-first/ second-order, Elovich, intraparticle diffusion), van't Hoff thermodynamics, batch fitting across many samples at once, FTIR baseline correction, automatic peak picking and functional-group analysis, XRD peak deconvolution and crystallinity index, BET surface area, TGA analysis (DTG curve with auto-detected decomposition peaks, moisture/volatile-matter/ash/fixed-carbon straight off a curve for a single sample or in batch across many, and Kissinger non-isothermal kinetics from multi-heating-rate data), proximate/ultimate analysis, and correlation matrices, without writing any R code. Results and 600 dpi TIFF figures can be downloaded directly from the browser, along with a combined analysis report.

This package provides a nonparametric, multicore-capable plausible naive Bayes classifier based on Pareto density estimation (PDE). It addresses low-evidence cases through a plausibility correction. To enhance the interpretability of the flexible naive Bayes classifier by revealing its posterior structure and feature-wise, class-specific evidence, posterior probabilities can be visualized as class-wise line plots for one-dimensional data or color-coded Voronoi diagrams for pairwise feature projections, and class-conditional PDE likelihoods as overlaid, mirrored density profiles resembling violin plots. Methodological details are provided by Stier, Q., Hoffmann, J. and Thrun, M. C. (2026) "Classifying with the Fine Structure of Distributions: Leveraging Distributional Information for Robust and Plausible Naive Bayes" <DOI:10.3390/make8010013>. For multicore computations, the implementation applies the general memory-sharing approach described by Thrun, M. C. and Märte, J. (2026) "memshare: Memory Sharing for Multicore Computation in R with an Application to Feature Selection by Mutual Information using PDE" <DOI:10.32614/RJ-2025-043>.

The proposed Generalized Autoregressive Conditional Heteroskedasticity (GARCH)-informed Long Short-Term Memory (LSTM) model follows the concept of physics-informed machine learning (PIML) by integrating established econometric knowledge of price volatility into a data-driven forecasting framework. In the model, conditional volatility estimated from the GARCH process is incorporated as an additional explanatory signal or volatility-based weighting component within the LSTM architecture. This enables the LSTM to learn nonlinear temporal dependencies while remaining informed by the underlying characteristics of agricultural price series, including volatility clustering, heteroscedasticity and market uncertainty. The optimized weighting parameter, lambda, controls the contribution of the GARCH-derived volatility information to the final prediction. Thus, the model combines the statistical interpretability of GARCH with the nonlinear learning capability of LSTM, producing a hybrid PIML framework that is more responsive to both normal price movements and periods of extreme market volatility. The methodology is motivated by hybrid forecasting framework proposed by Yeasin and Paul (2024) <doi:10.1007/s11227-023-05542-3>.

The 2-D spatial and temporal Epidemic Type Aftershock Sequence ('ETAS') Model is widely used to decluster earthquake data catalogs. Usually, the calculation of standard errors of the ETAS model parameter estimates is based on the Hessian matrix derived from the log-likelihood function of the fitted model. However, when an ETAS model is fitted to a local data set over a time period that is limited or short, the standard errors based on the Hessian matrix may be inaccurate. It follows that the asymptotic confidence intervals for parameters may not always be reliable. As an alternative, this package allows for the construction of bootstrap confidence intervals based on empirical quantiles for the parameters of the 2-D spatial and temporal ETAS model. This version improves on Version 0.1.0 of the package by enabling the study space window (renamed study region') to be polygonal rather than merely rectangular. A Japan earthquake data catalog is used in a second example to illustrate this new feature.

This package provides a unified method, called M statistic, is provided for detecting phylogenetic signals in continuous traits, discrete traits, and multi-trait combinations. Blomberg and Garland (2002) <doi:10.1046/j.1420-9101.2002.00472.x> provided a widely accepted statistical definition of the phylogenetic signal, which is the "tendency for related species to resemble each other more than they resemble species drawn at random from the tree". The M statistic strictly adheres to the definition of phylogenetic signal, formulating an index and developing a method of testing in strict accordance with the definition, instead of relying on correlation analysis or evolutionary models. The novel method equivalently expressed the textual definition of the phylogenetic signal as an inequality equation of the phylogenetic and trait distances and constructed the M statistic. The M statistic implemented in this package is based on the methodology described in Yao and Yuan (2025) <doi:10.1002/ece3.71106>. If you use this method in your research, please cite the paper.

This package provides tools for detecting, decomposing, and stress-testing temporal drift in repeated binary decision systems. Complements the decisionpaths package by shifting focus from path construction to system-level change over time. Implements five core analytic modules: (1) prevalence drift â did the overall decision rate change over time?; (2) transition drift â did the probability of switching or persisting change?; (3) entropy and stability trends â did path complexity evolve?; (4) group-differential drift â did the system drift differently across subgroups?; (5) change-point and regime-shift detection â did the system change abruptly after a policy or model update? Additionally provides a robustness module for testing stability of drift conclusions across analytic choices, and a sensitivity module for probing vulnerability to data problems including missingness, miscoding, and threshold shifts. Defines four original drift indices: the Decision Drift Index (DDI), Transition Drift Index (TDI), Group Differential Drift (GDD), and Cumulative Drift Burden (CDB). Applications include algorithmic audit, AI governance, education, health, and organisational research.

Free and open source fonts from Kreative Software:

Constructium is a fork of SIL Gentium designed specifically to support constructed scripts as encoded in the Under-ConScript Unicode Registry. It is ideal for mixed Latin, Greek, Cyrillic, IPA, and conlang text in web sites and documents.

Fairfax is a 6x12 bitmap font for terminals, text editors, IDEs, etc. It supports many scripts and a large number of Unicode blocks as well as constructed scripts as encoded in the Under-ConScript Unicode Registry, pseudographics and semigraphics, and tons of private use characters. It has been superceded by Fairfax HD but is still maintained.

Fairfax HD is a halfwidth scalable monospace font for terminals, text editors, IDEs, etc. It supports many scripts and a large number of Unicode blocks as well as constructed scripts as encoded in the Under-ConScript Unicode Registry, pseudographics and semigraphics, and tons of private use characters.

Kreative Square is a fullwidth scalable monospace font designed specifically to support pseudographics, semigraphics, and private use characters.

This package provides a comprehensive collection of statistical functions for climate change research. Provides tools for temporal trend detection based on the Mann-Kendall (MK) test (Mann 1945 <doi:10.2307/1907187>; Kendall 1975, ISBN:0852641990) and Sen's slope (Sen 1968 <doi:10.2307/2285891>), spatial autocorrelation using Moran's I (Moran 1950 <doi:10.2307/2332142>), extreme value analysis using the Generalised Extreme Value (GEV) distribution and Peaks-Over-Threshold (POT) method (Coles 2001 <doi:10.1007/978-1-4471-3675-0>), standardised drought indices including the Standardised Precipitation Index (SPI; McKee et al. 1993) and the Standardised Precipitation Evapotranspiration Index (SPEI; Vicente-Serrano et al. 2010 <doi:10.1175/2009JCLI2909.1>), and formal detection-attribution methods via optimal fingerprint regression and Empirical Orthogonal Function (EOF) analysis (Allen and Tett 1999 <doi:10.1007/s003820050291>), and apparent temperature via the heat index (Steadman 1979 <doi:10.1175/1520-0450(1979)018%3C0861:TAOSPI%3E2.0.CO;2>). Suitable for both station-level time series and gridded climate fields.

We provide full functionality to smooth L1 penalized regression operators and to compute regression estimates thereof. For this, the objective function of a user-specified regression operator is first smoothed using Nesterov smoothing (see Y. Nesterov (2005) <doi:10.1007/s10107-004-0552-5>), resulting in a modified objective function with explicit gradients everywhere. The smoothed objective function and its gradient are minimized via BFGS, and the obtained minimizer is returned. Using Nesterov smoothing, the smoothed objective function can be made arbitrarily close to the original (unsmoothed) one. In particular, the Nesterov approach has the advantage that it comes with explicit accuracy bounds, both on the L1/L2 difference of the unsmoothed to the smoothed objective functions as well as on their respective minimizers (see G. Hahn, S.M. Lutz, N. Laha, C. Lange (2020) <doi:10.1101/2020.09.17.301788>). A progressive smoothing approach is provided which iteratively smoothes the objective function, resulting in more stable regression estimates. A function to perform cross validation for selection of the regularization parameter is provided.

Lattice-based space-filling designs with fill or separation distance properties including interleaved lattice-based minimax distance designs proposed in Xu He (2017) <doi:10.1093/biomet/asx036>, interleaved lattice-based maximin distance designs proposed in Xu He (2018) <doi:10.1093/biomet/asy069>, interleaved lattice-based designs with low fill and high separation distance properties proposed in Xu He (2024) <doi:10.1137/23M156940X>, (sliced) rotated sphere packing designs proposed in Xu He (2017) <doi:10.1080/01621459.2016.1222289> and Xu He (2019) <doi:10.1080/00401706.2018.1458655>, densest packing-based maximum projections designs proposed in Xu He (2020) <doi:10.1093/biomet/asaa057> and Xu He (2018) <doi:10.48550/arXiv.1709.02062>, maximin distance designs for mixed continuous, ordinal, and binary variables proposed in Hui Lan and Xu He (2025) <doi:10.48550/arXiv.2507.23405>, and optimized and regularly repeated lattice-based Latin hypercube designs for large-scale computer experiments proposed in Xu He, Junpeng Gong, and Zhaohui Li (2025) <doi:10.48550/arXiv.2506.04582>.