Generalization of Shapiro-Wilk test for multivariate variables.

This package implements nonparametric bootstrap tests for detecting monotonicity in regression functions from Hall, P. and Heckman, N. (2000) <doi:10.1214/aos/1016120363> Includes tools for visualizing results using Nadaraya-Watson kernel regression and supports efficient computation with C++'. Tutorials and shiny application demo are available at <https://www.laylaparast.com/monotonicitytest> and <https://parastlab.shinyapps.io/MonotonicityTest>.

This package provides utility functions for multivariate analysis (factor analysis, discriminant analysis, and others). The package is primary written for the course Multivariate analysis and for the course Computer intensive methods at the masters program of Applied Statistics at University of Ljubljana.

Normally building a GODB is fairly complicated, involving downloading multiple database files and using these to build e.g. a mySQL database. Accessing this database is also complicated, involving an intimate knowledge of the database in order to construct reliable queries. Here we have a more modest goal, generating GOGOA3, which is a stripped down version of the GODB that was originally restricted to human genes as designated by the HUGO Gene Nomenclature Committee (HGNC) (see <https://geneontology.org/>). I have now added about two dozen additional species, namely all species represented on the Gene Ontology download page <https://current.geneontology.org/products/pages/downloads.html>. This covers most of the model organisms that are commonly used in bio-medical and basic research (assuming that anyone still has a grant to do such research). This can be built in a matter of seconds from 2 easily downloaded files (see <https://current.geneontology.org/products/pages/downloads.html> and <https://geneontology.org/docs/download-ontology/>), and it can be queried by e.g. w<-which(GOGOA3[,"HGNC"] %in% hgncList) where GOGOA3 is a matrix representing the minimalist GODB and hgncList is a list of gene identifiers. This database will be used in my upcoming package GoMiner which is based on my previous publication (see Zeeberg, B.R., Feng, W., Wang, G. et al. (2003)<doi:10.1186/gb-2003-4-4-r28>). Relevant .RData files are available from GitHub (<https://github.com/barryzee/GO/tree/main/databases>).

The MARSS package provides maximum-likelihood parameter estimation for constrained and unconstrained linear multivariate autoregressive state-space (MARSS) models, including partially deterministic models. MARSS models are a class of dynamic linear model (DLM) and vector autoregressive model (VAR) model. Fitting available via Expectation-Maximization (EM), BFGS (using optim), and TMB (using the marssTMB companion package). Functions are provided for parametric and innovations bootstrapping, Kalman filtering and smoothing, model selection criteria including bootstrap AICb, confidences intervals via the Hessian approximation or bootstrapping, and all conditional residual types. See the user guide for examples of dynamic factor analysis, dynamic linear models, outlier and shock detection, and multivariate AR-p models. Online workshops (lectures, eBook, and computer labs) at <https://atsa-es.github.io/>.

Build CPMs (cumulative probability models, also known as cumulative link models) to account for detection limits (both single and multiple detection limits) in response variables. Conditional quantiles and conditional CDFs can be calculated based on fitted models. The package implements methods described in Tian, Y., Li, C., Tu, S., James, N. T., Harrell, F. E., & Shepherd, B. E. (2022). "Addressing Detection Limits with Semiparametric Cumulative Probability Models". <arXiv:2207.02815>.

This package provides a Shiny application to estimate the sample size required for a metabolomic experiment to achieve a desired statistical power. Estimation is possible with or without available data from a pilot study.

Visualization of multi-dimensional data arising in multi-objective optimization, including plots of the empirical attainment function (EAF), M. López-Ibáñez, L. Paquete, and T. Stützle (2010) <doi:10.1007/978-3-642-02538-9_9>, and symmetric Vorob'ev expectation and deviation, M. Binois, D. Ginsbourger, O. Roustant (2015) <doi:10.1016/j.ejor.2014.07.032>, among others.

Efficient finite difference method for valuing European and American multi-asset options.

In many agricultural, engineering, industrial, post-harvest and processing experiments, the number of factor level changes and hence the total number of changes is of serious concern as such experiments may consists of hard-to-change factors where it is physically very difficult to change levels of some factors or sometime such experiments may require normalization time to obtain adequate operating condition. For this reason, run orders that offer the minimum number of factor level changes and at the same time minimize the possible influence of systematic trend effects on the experimentation have been sought. Factorial designs with minimum changes in factors level may be preferred for such situations as these minimally changed run orders will minimize the cost of the experiments. This technique can be employed to any half replicate of two level factorial run order where the number of factors are greater than two. For method details see, Bhowmik, A., Varghese, E., Jaggi, S. and Varghese, C. (2017) <doi:10.1080/03610926.2016.1152490>. This package generates all possible minimally changed two-level half-fractional factorial designs for different experimental setups along with various statistical criteria to measure the performance of these designs through a user-friendly interface. It consist of the function minimal.2halfFFD() which launches the application interface.

Extended tools for analyzing telemetry data using generalized hidden Markov models. Features of momentuHMM (pronounced ``momentum'') include data pre-processing and visualization, fitting HMMs to location and auxiliary biotelemetry or environmental data, biased and correlated random walk movement models, hierarchical HMMs, multiple imputation for incorporating location measurement error and missing data, user-specified design matrices and constraints for covariate modelling of parameters, random effects, decoding of the state process, visualization of fitted models, model checking and selection, and simulation. See McClintock and Michelot (2018) <doi:10.1111/2041-210X.12995>.

The MCC-F1 analysis is a method to evaluate the performance of binary classifications. The MCC-F1 curve is more reliable than the Receiver Operating Characteristic (ROC) curve and the Precision-Recall (PR)curve under imbalanced ground truth. The MCC-F1 analysis also provides the MCC-F1 metric that integrates classifier performance over varying thresholds, and the best threshold of binary classification.

The implemented methods reach out to scientists that seek to estimate multiplicity of infection (MOI) and lineage (allele) frequencies and prevalences at molecular markers using the maximum-likelihood method described in Schneider (2018) <doi:10.1371/journal.pone.0194148>, and Schneider and Escalante (2014) <doi:10.1371/journal.pone.0097899>. Users can import data from Excel files in various formats, and perform maximum-likelihood estimation on the imported data by the package's moimle() function.

This package provides samplers for various matrix variate distributions: Wishart, inverse-Wishart, normal, t, inverted-t, Beta type I, Beta type II, Gamma, confluent hypergeometric. Allows to simulate the noncentral Wishart distribution without the integer restriction on the degrees of freedom.

Extends the mlr3 ML framework with methods for spatial objects. Data storage and prediction are supported for packages terra', raster and stars'.

The ultimate goal is to support 2-2-1, 2-1-1, and 1-1-1 models for multilevel mediation, the option of a moderating variable for either the a, b, or both paths, and covariates. Currently the 1-1-1 model is supported and several options of random effects; the initial code for bootstrapping was evaluated in simulations by Falk, Vogel, Hammami, and MioÄ eviÄ (2024) <doi:10.3758/s13428-023-02079-4>. Support for Bayesian estimation using brms comprises ongoing work. Currently only continuous mediators and outcomes are supported. Factors for any predictors must be numerically represented.

Lattice functions for drawing folded empirical cumulative distribution plots, or mountain plots. A mountain plot is similar to an empirical CDF plot, except that the curve increases from 0 to 0.5, then decreases from 0.5 to 1 using an inverted scale at the right side. See Monti (1995) <doi:10.1080/00031305.1995.10476179>.

Electronic health records (EHR) linked with biorepositories are a powerful platform for translational studies. A major bottleneck exists in the ability to phenotype patients accurately and efficiently. Towards that end, we developed an automated high-throughput phenotyping method integrating International Classification of Diseases (ICD) codes and narrative data extracted using natural language processing (NLP). Specifically, our proposed method, called MAP (Map Automated Phenotyping algorithm), fits an ensemble of latent mixture models on aggregated ICD and NLP counts along with healthcare utilization. The MAP algorithm yields a predicted probability of phenotype for each patient and a threshold for classifying subjects with phenotype yes/no (See Katherine P. Liao, et al. (2019) <doi:10.1093/jamia/ocz066>.).

This package implements the MST-kNN clustering algorithm which was proposed by Inostroza-Ponta, M. (2008) <https://trove.nla.gov.au/work/28729389?selectedversion=NBD44634158>.

This package provides a complete toolkit to process the Munich ChronoType Questionnaire (MCTQ) for its three versions (standard, micro, and shift). MCTQ is a quantitative and validated tool to assess chronotypes using peoples sleep behavior, originally presented by Till Roenneberg, Anna Wirz-Justice, and Martha Merrow (2003, <doi:10.1177/0748730402239679>).

The multiple instance data set consists of many independent subjects (called bags) and each subject is composed of several components (called instances). The outcomes of such data set are binary or categorical responses, and, we can only observe the subject-level outcomes. For example, in manufacturing processes, a subject is labeled as "defective" if at least one of its own components is defective, and otherwise, is labeled as "non-defective". The milr package focuses on the predictive model for the multiple instance data set with binary outcomes and performs the maximum likelihood estimation with the Expectation-Maximization algorithm under the framework of logistic regression. Moreover, the LASSO penalty is attached to the likelihood function for simultaneous parameter estimation and variable selection.

This package provides a simple and effective tool for computing and visualizing statistical power for meta-analysis, including power analysis of main effects (Jackson & Turner, 2017)<doi:10.1002/jrsm.1240>, test of homogeneity (Pigott, 2012)<doi:10.1007/978-1-4614-2278-5>, subgroup analysis, and categorical moderator analysis (Hedges & Pigott, 2004)<doi:10.1037/1082-989X.9.4.426>.

Addons for the mice package to perform multiple imputation using chained equations with two-level data. Includes imputation methods dedicated to sporadically and systematically missing values. Imputation of continuous, binary or count variables are available. Following the recommendations of Audigier, V. et al (2018) <doi:10.1214/18-STS646>, the choice of the imputation method for each variable can be facilitated by a default choice tuned according to the structure of the incomplete dataset. Allows parallel calculation and overimputation for mice'.

To determine the number of quantitative assays needed for a sample of data using pooled testing methods, which include mini-pooling (MP), MP with algorithm (MPA), and marker-assisted MPA (mMPA). To estimate the number of assays needed, the package also provides a tool to conduct Monte Carlo (MC) to simulate different orders in which the sample would be collected to form pools. Using MC avoids the dependence of the estimated number of assays on any specific ordering of the samples to form pools.