Efficient variational inference methods for fully Bayesian univariate and multivariate Gaussian and t-process regression models. Hierarchical shrinkage priors, including the triple gamma prior, are used for effective variable selection and covariance shrinkage in high-dimensional settings. The package leverages normalizing flows to approximate complex posterior distributions. For details on implementation, see Knaus (2025) <doi:10.48550/arXiv.2501.13173>.

This package provides a set of functions to quantify and visualise social autocorrelation.

Analysis of species count data in ecology often requires normalization to an identical sample size. Rarefying (random subsampling without replacement), which is a popular method for normalization, has been widely criticized for its poor reproducibility and potential distortion of the community structure. In the context of microbiome count data, researchers explicitly advised against the use of rarefying. An alternative to rarefying is scaling with ranked subsampling (SRS). SRS consists of two steps. In the first step, the total counts for all OTUs (operational taxonomic units) or species in each sample are divided by a scaling factor chosen in such a way that the sum of the scaled counts Cscaled equals Cmin. In the second step, the non-integer Cscaled values are converted into integers by an algorithm that we dub ranked subsampling. The Cscaled value for each OTU or species is split into the integer part Cint (Cint = floor(Cscaled)) and the fractional part Cfrac (Cfrac = Cscaled - Cints). Since the sum of Cint is smaller or equal to Cmin, additional delta C = Cmin - the sum of Cint counts have to be added to the library to reach the total count of Cmin. This is achieved as follows. OTUs are ranked in the descending order of their Cfrac values. Beginning with the OTU of the highest rank, single count per OTU is added to the normalized library until the total number of added counts reaches delta C and the sum of all counts in the normalized library equals Cmin. When the lowest Cfrag involved in picking delta C counts is shared by several OTUs, the OTUs used for adding a single count to the library are selected in the order of their Cint values. This selection minimizes the effect of normalization on the relative frequencies of OTUs. OTUs with identical Cfrag as well as Cint are sampled randomly without replacement. See Beule & Karlovsky (2020) <doi:10.7717/peerj.9593> for details.

This package provides comprehensive analytics, reporting, and testing capabilities for systematic review search strategies. The package focuses on validating search performance, generating standardized PRISMA'-compliant reports, and ensuring reproducibility in evidence synthesis. Features include precision-recall analysis, cross-database performance comparison, benchmark validation against gold standards, sensitivity analysis, temporal coverage assessment, automated report generation, and statistical comparison of search strategies. Supports multiple export formats including CSV', Excel', RIS', BibTeX', and EndNote'. Includes tools for duplicate detection, search strategy optimization, cross-validation frameworks, meta-analysis of benchmark results, power analysis for study design, and reproducibility package creation. Optionally connects to PubMed for direct database searching and real-time strategy comparison using the E-utilities API'. Enhanced with bootstrap comparison methods, McNemar test for strategy evaluation, and comprehensive visualization tools for performance assessment. Methods based on Manning et al. (2008) for information retrieval metrics, Moher et al. (2009) for PRISMA guidelines, and Sampson et al. (2006) for systematic review search methodology.

Hail is an open-source, general-purpose, python based data analysis tool with additional data types and methods for working with genomic data, see <https://hail.is/>. Hail is built to scale and has first-class support for multi-dimensional structured data, like the genomic data in a genome-wide association study (GWAS). Hail is exposed as a python library, using primitives for distributed queries and linear algebra implemented in scala', spark', and increasingly C++'. The sparkhail is an R extension using sparklyr package. The idea is to help R users to use hail functionalities with the well-know tidyverse syntax, see <https://www.tidyverse.org/>.

Inter-rater reliability analysis for binary classification tasks involving two or more raters within a signal detection-theoretic framework. User-supplied rating data are standardised into a common long-format structure. The package automatically computes Cohen's kappa for two raters and Fleiss kappa for multiple raters. When ground-truth labels are available, rater-specific hit rates, false-alarm rates, sensitivity, specificity, and decision thresholds are estimated from observed classification responses using standard signal detection-theoretic transformations (DeCarlo, 1998) <doi:10.1037/1082-989X.3.2.186>. The package implements the Strategic Convergence Index (SCI; Gianeselli, 2026) <doi:10.1177/00131644261417643>, defined as SCI = 1 - [Var(t_i) / Var_max], where Var(t_i) denotes the variance of rater-specific decision thresholds and Var_max denotes the reference variance under maximal threshold dispersion. SCI quantifies convergence in rater decision criteria beyond observed agreement alone and complements classical agreement coefficients by distinguishing agreement in observed categorical outcomes from convergence in latent decision thresholds under an explicit signal detection-theoretic model of categorical judgment. The package provides structured summaries and threshold-based diagnostics for applications in which similar agreement coefficients may reflect substantively different underlying decision criteria across raters.

This app enables interactive validation, interpretation and visualization of structural topic models from the stm package by Roberts and others (2014) . It also includes helper functions for model diagnostics and extracting data from effect estimates.

This package implements methods for variable selection in linear regression based on the "Sum of Single Effects" (SuSiE) model, as described in Wang et al (2020) <DOI:10.1101/501114> and Zou et al (2021) <DOI:10.1101/2021.11.03.467167>. These methods provide simple summaries, called "Credible Sets", for accurately quantifying uncertainty in which variables should be selected. The methods are motivated by genetic fine-mapping applications, and are particularly well-suited to settings where variables are highly correlated and detectable effects are sparse. The fitting algorithm, a Bayesian analogue of stepwise selection methods called "Iterative Bayesian Stepwise Selection" (IBSS), is simple and fast, allowing the SuSiE model be fit to large data sets (thousands of samples and hundreds of thousands of variables).

R version of scperturb tool for single-cell perturbation analysis. Contains wrappers for performing E-statistics for Seurat objects. More details on the method can be found in Peidli et al. (2023) <doi:10.1101/2022.08.20.504663> and in Székely and Rizzo (2004).

This package provides a pipeline to perform small area estimation and prevalence mapping of binary indicators using health and demographic survey data, described in Dong et al. (2026) <doi:10.1093/jssam/smaf048>, Wakefield et al. (2025) <doi:10.48550/arXiv.2110.09576> and Wakefield et al. (2020) <doi:10.1111/insr.12400>.

This package provides a consistent, semi-supervised, non-parametric survival curve estimator optimized for efficient use of Electronic Health Record (EHR) data with a limited number of current status labels. See van der Laan and Robins (1997) <doi:10.2307/2670119>.

Convert text (and text in R objects) to Mocking SpongeBob case <https://knowyourmeme.com/memes/mocking-spongebob> and show them off in fun ways. CoNVErT TexT (AnD TeXt In r ObJeCtS) To MOCkINg SpoNgebOb CAsE <https://knowyourmeme.com/memes/mocking-spongebob> aND shOw tHem OFf IN Fun WayS.

Parameter inference methods for models defined implicitly using a random simulator. Inference is carried out using simulation-based estimates of the log-likelihood of the data. The inference methods implemented in this package are explained in Park, J. (2025) <doi:10.48550/arxiv.2311.09446>. These methods are built on a simulation metamodel which assumes that the estimates of the log-likelihood are approximately normally distributed with the mean function that is locally quadratic around its maximum. Parameter estimation and uncertainty quantification can be carried out using the ht() function (for hypothesis testing) and the ci() function (for constructing a confidence interval for one-dimensional parameters).

The nature of working with structured query language ('SQL') scripts efficiently often requires the creation of temporary tables and there are few clean and simple R SQL execution approaches that allow you to complete this kind of work with the R environment. This package seeks to give SQL implementations in R a little love by deploying functions that allow you to deploy complex SQL scripts within a typical R workflow.

This package provides a statistical method for reducing the number of covariates in an analysis by evaluating Variable Importance Measures (VIMPs) derived from the Random Forest algorithm. It performs statistical tests on the VIMPs and outputs whether the covariate is significant along with the p-values.

Identifying spatially variable genes is critical in linking molecular cell functions with tissue phenotypes. This package utilizes a granularity-based dimension-agnostic tool, single-cell big-small patch (scBSP), implementing sparse matrix operation and KD tree methods for distance calculation, for the identification of spatially variable genes on large-scale data. The detailed description of this method is available at Wang, J. and Li, J. et al. 2023 (Wang, J. and Li, J. (2023), <doi:10.1038/s41467-023-43256-5>).

Collection (syllogi in greek) of real and fictitious data sets for teaching purposes. The datasets were manually entered by the author from the respective references as listed in the individual dataset documentation. The fictions datasets are the creation of the author, that he has found useful for teaching statistics.

Convert a time series of observations to a time series of standardised indices that can be used to monitor variables on a common and probabilistically interpretable scale. The indices can be aggregated and rescaled to different time scales, visualised using plot capabilities, and calculated using a range of distributions. This includes flexible non-parametric and non-stationary methods.

Allows objects to be stored on disc and automatically recalled into memory, as required, by delayed assignment.

These are tools that allow users to do time series diagnostics, primarily tests of unit root, by way of simulation. While there is nothing necessarily wrong with the received wisdom of critical values generated decades ago, simulation provides its own perks. Not only is simulation broadly informative as to what these various test statistics do and what are their plausible values, simulation provides more flexibility for assessing unit root by way of different thresholds or different hypothesized distributions.

All data in the book "Statistical Methods in Biology" by Welham et al. (2015) <doi:10.1201/b17336> with a corresponding documentation and illustrative analysis of the data.

Simulate and plot general experimental crosses. The focus is on simulating genotypes with an aim towards flexibility rather than speed. Meiosis is simulated following the Stahl model, in which chiasma locations are the superposition of two processes: a proportion p coming from a process exhibiting no interference, and the remainder coming from a process following the chi-square model.

Allows the user to estimate a vector logistic smooth transition autoregressive model via maximum log-likelihood or nonlinear least squares. It further permits to test for linearity in the multivariate framework against a vector logistic smooth transition autoregressive model with a single transition variable. The estimation method is discussed in Terasvirta and Yang (2014, <doi:10.1108/S0731-9053(2013)0000031008>). Also, realized covariances can be constructed from stock market prices or returns, as explained in Andersen et al. (2001, <doi:10.1016/S0304-405X(01)00055-1>).

There is variation across AgNPs due to differences in characterization techniques and testing metrics employed in studies. To address this problem, we have developed a systematic evaluation framework called sysAgNPs'. Within this framework, Distribution Entropy (DE) is utilized to measure the uncertainty of feature categories of AgNPs, Proclivity Entropy (PE) assesses the preference of these categories, and Combination Entropy (CE) quantifies the uncertainty of feature combinations of AgNPs. Additionally, a Markov chain model is employed to examine the relationships among the sub-features of AgNPs and to determine a Transition Score (TS) scoring standard that is based on steady-state probabilities. The sysAgNPs framework provides metrics for evaluating AgNPs, which helps to unravel their complexity and facilitates effective comparisons among different AgNPs, thereby advancing the scientific research and application of these AgNPs.