Correlation coefficients for multivariate data, namely the squared correlation coefficient and the RV coefficient (multivariate generalization of the squared Pearson correlation coefficient). References include Mardia K.V., Kent J.T. and Bibby J.M. (1979). "Multivariate Analysis". ISBN: 978-0124712522. London: Academic Press.

Generates replicated sets of sequences with Monte Carlo simulated timing changes and computes various indicators for evaluating effects of timing uncertainty on sequence analysis results. See Ritschard, G. and Liao, T.F. (2026): "Assessing the Impact of Timing Errors in Sequence Analysis". International Journal of Social Research Methodology <doi:10.1080/13645579.2026.2666297>.

Allows the user to estimate transition probabilities for migratory animals between any two phases of the annual cycle, using a variety of different data types. Also quantifies the strength of migratory connectivity (MC), a standardized metric to quantify the extent to which populations co-occur between two phases of the annual cycle. Includes functions to estimate MC and the more traditional metric of migratory connectivity strength (Mantel correlation) incorporating uncertainty from multiple sources of sampling error. For cross-species comparisons, methods are provided to estimate differences in migratory connectivity strength, incorporating uncertainty. See Cohen et al. (2018) <doi:10.1111/2041-210X.12916>, Cohen et al. (2019) <doi:10.1111/ecog.03974>, Roberts et al. (2023) <doi:10.1002/eap.2788>, and Hostetler et al. (2025) <doi:10.1111/2041-210X.14467> for details on some of these methods.

Partial Replacement Imputation Estimation (PRIME) can overcome problems caused by missing covariates in additive partially linear model. PRIME conducts imputation and regression simultaneously with known and unknown model structure. More details can be referred to Zishu Zhan, Xiangjie Li and Jingxiao Zhang. (2022) <arXiv:2205.14994>.

The target of margaret is help to extract data from Minciencias to analyze scientific production in Colombia.

Nonparametric estimation and inference for natural direct and indirect effects by Chan, Imai, Yam and Zhang (2016) <arXiv:1601.03501>.

Forecasting of univariate time series using feature extraction with variable prediction methods is provided. Feature extraction is done with a redundant Haar wavelet transform with filter h = (0.5, 0.5). The advantage of the approach compared to typical Fourier based methods is an dynamic adaptation to varying seasonalities. Currently implemented prediction methods based on the selected wavelets levels and scales are a regression and a multi-layer perceptron. Forecasts can be computed for horizon 1 or higher. Model selection is performed with an evolutionary optimization. Selection criteria are currently the AIC criterion, the Mean Absolute Error or the Mean Root Error. The data is split into three parts for model selection: Training, test, and evaluation dataset. The training data is for computing the weights of a parameter set. The test data is for choosing the best parameter set. The evaluation data is for assessing the forecast performance of the best parameter set on new data unknown to the model. This work is published in Stier, Q.; Gehlert, T.; Thrun, M.C. Multiresolution Forecasting for Industrial Applications. Processes 2021, 9, 1697. <doi:10.3390/pr9101697>.

This package implements an MCMC sampler for the posterior distribution of arbitrary time-homogeneous multivariate stochastic differential equation (SDE) models with possibly latent components. The package provides a simple entry point to integrate user-defined models directly with the sampler's C++ code, and parallelizes large portions of the calculations when compiled with OpenMP'.

Engaging the private sector in contraceptive method supply is critical for equitable, sustainable, and accessible healthcare systems. This package implements Bayesian hierarchical models to estimate public and private contraceptive supply shares over time at national and subnational levels, using Demographic and Health Survey (DHS) data. Penalized splines are used to track supply shares over time, and spatial correlation structures link national and subnational estimates in data- sparse settings. For more details see Comiskey (2025) <doi:10.48550/arXiv.2510.25153>.

The time series forecasting framework for use with the tidymodels ecosystem. Models include ARIMA, Exponential Smoothing, and additional time series models from the forecast and prophet packages. Refer to "Forecasting Principles & Practice, Second edition" (<https://otexts.com/fpp2/>). Refer to "Prophet: forecasting at scale" (<https://research.facebook.com/blog/2017/02/prophet-forecasting-at-scale/>.).

This package provides a flexible framework for power analysis using Monte Carlo simulation for settings in which considerations of the correlations between predictors are important. Users can set up a data generative model that preserves dependence structures among predictors given existing data (continuous, binary, or ordinal). Users can also generate power curves to assess the trade-offs between sample size, effect size, and power of a design. This package includes several statistical models common in environmental mixtures studies. For more details and tutorials, see Nguyen et al. (2022) <arXiv:2209.08036>.

Classify missing data as missing completely at random (MCAR), missing at random (MAR), or missing not at random (MNAR). This step is required before handling missing data (e.g. mean imputation) so that bias is not introduced. See Little (1988) <doi:10.1080/01621459.1988.10478722> for the statistical rationale for the methods used.

This package provides a simple and trustworthy methodology for the analysis of misreported continuous time series using either a frequentist (bootstrap-based EM algorithm) or a Bayesian (MCMC via JAGS) approach. The frequentist method is described in Morina et al. (2021) <doi:10.1038/s41598-021-02620-5>. The Bayesian extension fits the same ARMA model with misreporting structure using a full posterior distribution, providing credible intervals and DIC for model comparison, as described in Morina et al. (2024) <doi:10.1101/2024.02.26.24303373>.

Assessment of inconsistency in meta-analysis by calculating the Decision Inconsistency index (DI) and the Across-Studies Inconsistency (ASI) index. These indices quantify inconsistency taking into account outcome-level decision thresholds.

The microplot function writes a set of R graphics files to be used as microplots (sparklines) in tables in either LaTeX', HTML', Word', or Excel files. For LaTeX', we provide methods for the Hmisc::latex() generic function to construct latex tabular environments which include the graphs. These can be used directly with the operating system pdflatex or latex command, or by using one of Sweave', knitr', rmarkdown', or Emacs org-mode as an intermediary. For MS Word', the msWord() function uses the flextable package to construct Word tables which include the graphs. There are several distinct approaches for constructing HTML files. The simplest is to use the msWord() function with argument filetype="html". Alternatively, use either Emacs org-mode or the htmlTable::htmlTable() function to construct an HTML file containing tables which include the graphs. See the documentation for our as.htmlimg() function. For Excel use on Windows', the file examples/irisExcel.xls includes VBA code which brings the individual panels into individual cells in the spreadsheet. Examples in the examples and demo subdirectories are shown with lattice graphics, ggplot2 graphics, and base graphics. Examples for LaTeX include Sweave (both LaTeX'-style and Noweb'-style), knitr', emacs org-mode', and rmarkdown input files and their pdf output files. Examples for HTML include org-mode and Rmd input files and their webarchive HTML output files. In addition, the as.orgtable() function can display a data.frame in an org-mode document. The examples for MS Word (with either filetype="docx" or filetype="html") work with all operating systems. The package does not require the installation of LaTeX or MS Word to be able to write .tex or .docx files.

This package implements contamination bias diagnostics and alternative estimators for regressions with multiple treatments. The implementation is based on Goldsmith-Pinkham, Hull, and Kolesár (2024) <doi:10.48550/arXiv.2106.05024>.

With the deprecation of mocking capabilities shipped with testthat as of edition 3 it is left to third-party packages to replace this functionality, which in some test-scenarios is essential in order to run unit tests in limited environments (such as no Internet connection). Mocking in this setting means temporarily substituting a function with a stub that acts in some sense like the original function (for example by serving a HTTP response that has been cached as a file). The only exported function with_mock() is modeled after the eponymous testthat function with the intention of providing a drop-in replacement.

An implementation of matrix mathematics wherein operations are performed "by name.".

This package provides methods for maximum likelihood and Bayesian estimation for the Wishart mixture model and the mixture-of-experts Wishart (MoE-Wishart) model. The package provides four inference algorithms for these models, each implemented using the expectationâ maximization (EM) algorithm for maximum likelihood estimation and a fully Bayesian approach via Gibbs-within-Metropolisâ Hastings sampling.

An implementation of popular screening methods that are commonly employed in ultra-high and high dimensional data. Through this publicly available package, we provide a unified framework to carry out model-free screening procedures including SIS (Fan and Lv (2008) <doi:10.1111/j.1467-9868.2008.00674.x>), SIRS(Zhu et al. (2011)<doi:10.1198/jasa.2011.tm10563>), DC-SIS (Li et al. (2012) <doi:10.1080/01621459.2012.695654>), MDC-SIS(Shao and Zhang (2014) <doi:10.1080/01621459.2014.887012>), Bcor-SIS (Pan et al. (2019) <doi:10.1080/01621459.2018.1462709>), PC-Screen (Liu et al. (2020) <doi:10.1080/01621459.2020.1783274>), WLS (Zhong et al.(2021) <doi:10.1080/01621459.2021.1918554>), Kfilter (Mai and Zou (2015) <doi:10.1214/14-AOS1303>), MVSIS (Cui et al. (2015) <doi:10.1080/01621459.2014.920256>), PSIS (Pan et al. (2016) <doi:10.1080/01621459.2014.998760>), CAS (Xie et al. (2020) <doi:10.1080/01621459.2019.1573734>), CI-SIS (Cheng and Wang. (2023) <doi:10.1016/j.cmpb.2022.107269>), CSIS (Cheng et al. (2024) <doi:10.1007/s00180-023-01399-5>) and Log-rank SIS.

This package provides a general framework for computing powers of matrices. A key feature is the capability for users to write callback functions, called after each iteration, thus enabling customization for specific applications. Diverse types of matrix classes/matrix multiplication are accommodated. If the multiplication type computes in parallel, then the package computation is also parallel.

This package performs treatment assignment for (field) experiments considering available, possibly multivariate and continuous, information (covariates, observable characteristics), that is: forms balanced treatment groups, according to the minMSE-method as proposed by Schneider and Schlather (2017) <DOI:10419/161931>.

This package implements the method of successive dichotomizations by Bradley and Massof (2018) <doi:10.1371/journal.pone.0206106>, which estimates item measures, person measures and ordered rating category thresholds given ordinal rating scale data.

Integrates large language model generated item responses into psychometric calibration studies through a mixed-subjects design for unidimensional two-parameter and one-parameter logistic item response theory models. Human pilot responses are augmented with model-generated responses using a prediction-powered inference estimator (Angelopoulos, Bates, Fannjiang, Jordan and Zrnic (2023) <doi:10.1126/science.adi6000>; Angelopoulos, Duchi and Zrnic (2023) <doi:10.48550/arXiv.2311.01453>) adapted to marginal maximum-likelihood estimation, following the mixed-subjects design of Broska, Howes and van Loon (2025) <doi:10.1177/00491241251326865>. The estimator is anchored to the human responses and is asymptotically unbiased for the human item parameters at any tuning weight; the weight on the synthetic responses is chosen to minimize propagated ability-score risk, down-weighting uninformative or biased generated responses. Louis-corrected sandwich standard errors, ability scoring, cross-fitted tuning, and scale linking are also provided.