Two hundred eleven new packages made it to CRAN in November. Here are my “Top 40” picks in thirteen categories: Computational Methods, Data, Ecology, Finance, Genomics, Humanities, Machine Learning, Medicine, Networks, Statistics, Time Series, Utilities, and Visualization. It was gratifying to see multiple packages developed for applications in the Computational Humanities. R is helping to extend the reach of data literacy.
bigQF v1.6: Implements a computationally efficient leading-eigenvalue approximation to tail probabilities and quantiles of large quadratic forms (e.g. in the sequence kernel association test) and also provides stochastic singular value decomposition for dense or sparse matrices. See Lumley et al. (2018) for background and the vignettes: Checking pQF vs SKAT, Matrix-free computations, SKAT, weights, and projections.
japanstat v0.1.0: Provides tools for using the API of e-Stat, a portal site for Japanese government statistics which include functions for automatic query generation, data collection and formatting. See README for examples.
mFD v1.0.1: Provides functions to compute functional traits-based distances between pairs of species for species gathered in assemblages. Chao et al. (2018) and Mouillot et al. (2013) along with the other references listed for background. There are five vignettes including General Workflow and Compute and Interpret Quality of Functional Spaces
popbayes v1.0: Implements a Bayesian framework to infer the trends of animal populations over time from series of counts by accounting for count precision, smoothing the population rate of increase over time, and accounting for the maximum demographic potential of species. There is a Get started vignette and another on Getting species information.
MultiATSM v0.0.1: Provides functions to model the affine term structure of interest rates based on the single-country unspanned macroeconomic risk framework of Joslin et al. (2014) and the multi country extensions of Jotikasthira et al.(2015), and Candelon and Moura (2021). See the vignette for an example.
phylosamp v0.1.6: Implements tools to estimate the probability of true transmission between two cases given phylogenetic linkage and the expected number of true transmission links in a sample. The methods are described in Wohl et al. (2021). There are vignettes on estimating FDR from sample size, Sample size from FDR, Sensitivity and specificity, and Examples.
Rtropical v1.2.1: Provides functions to process phylogenetic trees with tropical support vector machines and principal component analysis defined with tropical geometry. See Tang et al. (2020) for details about tropical SVMs, Page et al. (2020) and Yoshida et al. (2019) for tropical PCA, and Ardila & Develin (2007) for some background on tropical mathematics. The vignette will get you started.
litRiddle v0.4.1: Provides a dataset, functions to explore the quality of literary novels, the data of a reader survey about fiction in Dutch, a description of the novels the readers rated, and the results of stylistic measurements of the novels. There is a vignette.
kairos v1.0.0: Provides a toolkit for absolute dating and analysis of chronological patterns, including functions for chronological modeling and dating of archaeological assemblages from count data. There is a Manual, and a Bibliography.
innsight v0.1.1: Implements methods to analyze the behavior and individual predictions of modern neural networks including Connection Weights as described by Olden et al. (2004), Layer-wise Relevance Propagation as described by Bach et al. (2015), Deep Learning Important Features as described by Shrikumar et al. (2017), and gradient-based methods like SmoothGrad described by Smilkov et al. (2017), and Gradient x Input described by Baehrens et al. (2009). There is an Introduction and a vignette on Custom Model Definition.
topicmodels.etm v0.1.0: Provides functions to find topics in texts which are semantically embedded using techniques like word2vec or GloVe. See Dieng et al. (2019) for the details, and look here for an example.
eSIR v0.4.2: Implements the extended state-space SIR models developed by Song Lab at UM school of Public Health which include capabilities to model time-varying transmission, time-dependent quarantine, and time-dependent antibody-immunization. See Wang et al. (2020) for background, and README for examples.
QHScrnomo v2.2.0: Provides functions for fitting and predicting competing risk models, creating nomograms, k-fold cross validation, calculating the discrimination metric, and drawing calibration curves. See the vignette for a short tutorial.
rnmamod v0.1.0: Implements a comprehensive suite of functions to perform and visualize pairwise and network meta-analyses. The package covers core Bayesian one-stage models implemented in a systematic review with multiple interventions, including fixed-effect and random-effects network meta-analysis. There are vignettes on Performing a network meta-analysis and Describing the network.
whomds v1.0.1: Provides functions for calculating and presenting the results from the WHO Model Disability Survey (MDA). See Andrich (2014) for background on Rasch measurement. There are English and Spanish versions of six vignettes including Background on disability measurement and Best practices with Rasch Analysis.
diffudist v1.0.0: Enables the evaluation of diffusion distances for complex single-layer networks by providing functions to evaluate the Laplacians, stochastic matrices, and the corresponding diffusion distance matrices. See De Domenico (2017) and Bertagnolli & De Domenico (2021) for the details and the vignette for some theory and examples.
NetFACS V0.3.0: Provides functions to analyze and visualize facial communication data, based on network theory and resampling methods and primarily targeted at datasets of facial expressions coded with the Facial Action Coding System. See Farine (2017) and Carsey & Harden (2014) for background, and the vignette for an introduction.
OneSampleMR v0.1.0: Provides functions for one sample Mendelian randomization and instrumental variable analyses including implementations of the Sanderson & Windmeijer (2016) conditional F-statistic, the multiplicative structural mean model Hernán and Robins (2006), and two-stage predictor substitution and two-stage residual inclusion estimators explained by Terza et al. (2008). There are short vignettes on the Multiplicative structural mean model and the Coditional F-Statistic.
PhaseTypeR v1.0.1: Implements functions to model continuous and discrete phase-type distributions, both univariate and multivariate. See Navarro (2019) for background. There is an Introduction and vignettes on Population Genetics and The Site Frequency Spectrum.
stan4bart v0.0-2: Fits semiparametric linear and multilevel models with non-parametric additive Bayesian additive regression trees BART and
Stan. Multilevel models can be expressed using
lme4 syntax. Look here to get started.
zoid v1.0.0: Fits Dirichlet regression and zero-and-one inflated Dirichlet regression (also called trinomial mixture models) with Bayesian methods implemented in
Stan. There are vignettes on Fitting Models, Simulating data, Priors, and Prior sensitivity for overdispersion.
surveil v0.1.0: Fits time series models for routine disease surveillance tasks and returns probability distributions for a variety of quantities of interest, including measures of health inequality, period and cumulative percent change, and age-standardized rates. See the vignette for an introduction.
VisitorCounts v1.0.0: Provides functions for modeling and forecasting of park visitor counts using social media data and (partial) on-site visitor counts. See Wood et al. (2013) for background and the vignette for examples.
dateutils v0.1.5: Provides utilities for mixed frequency data, in particular, to aggregate and normalize tabular mixed frequency data, index dates to end of period, and seasonally adjust tabular data. There is an Introduction.
lambdar v1.1.0: Provides functions for serving containers that can execute R code on the AWS Lambda serverless compute service. There are five vignettes including api-gateway-invocations and lambda-runtime-in-container.
mtb v0.1.1: Provides functions to summarize time related data, generate axis transformation from data, and assist Markdown and Shiny file editing. There are vignettes on working with axes, color, Markdown, and time related data.
rocker v0.2.1: Provides an
R6 class interface for handling database connections using the
DBI package as backend which allows handling of connections to
SQLite and other databases. There are seven short vignettes including DBI package objects and functions in R6 rocker class and Database transactions with R6 rocker class.
NHSRplotthedots v0.1.0: Provides tools for drawing statistical process control charts with functions to draw XmR charts, use change points and apply rules with summary indicators for when rules are breached. There is an Introduction and vignettes Deviations from Excel defaults and Number of points required.
shinymodels v0.1.0: Allows users to launch a
shiny application for
tidymodels results. For classification or regression models, the app can be used to determine if there is lack of fit or poorly predicted points. Look here to get started.