By my count, two hundred and thirty-three packages stuck to CRAN last month. I have tried to capture something of the diversity of the offerings by selecting packages in ten categories: Computational Methods, Data, Machine Learning, Medicine, Science, Shiny, Statistics, Time Series, Utilities, and Visualization. The Shiny category contains packages that expand on Shiny capabilities, not just packages that implement a Shiny application. It is not clear whether this is going to be a new cottage industry or not.
latte v0.2.1: Implements connections to
LattE for counting lattice points and integration inside convex polytopes, and
4ti2 for algebraic, geometric, and combinatorial problems on linear spaces and front-end tools facilitating their use in the ‘R’ ecosystem. Look here for an example.
nvctr v0.1.1: Implements the n-vector approach to calculating geographical positions using an ellipsoidal model of the Earth. This package is a translation of the FFi
Matlab library from FFI described in Gade (2010). The vignette provides examples.
EHRtemporalVariability v1.0: Provides functions to delineate reference changes over time in Electronic Health Records through the projection and visualization of dissimilarities among data temporal batches, and explore results through data temporal heat maps, information geometric temporal (IGT) plots, and a Shiny app. The vignette shows how to use the package.
kayadata v0.4.0: Provides data for Kaya identity variables (population, gross domestic product, primary energy consumption, and energy-related CO2 emissions), and includes utility functions for exploring and plotting fuel mix for a given country or region. See the vignette for examples.
raustats v0.1.0: Provides functions for downloading Australian economic statistics from the Australian Bureau of Statistics and Reserve Bank of Australia websites. The vignette shows how to use the package.
akmedoids v0.1.2: Advances a set of R-functions for longitudinal clustering of long-term trajectories, and determines the optimal solution based on the Caliński-Harabasz criterion ( Caliński and Harabasz (1974) ). The vignette works through an extended example.
shapper v0.1.0: Implements a wrapper for the Python
shap library that provides SHapley Additive exPlanations (SHAP) for the variables that influence particular observations in machine learning models. There are vignettes for classification and regression.
xgb2sql v0.1.2: Enables in-database scoring of
XGBoost models built in R, by translating trained model objects into SQL query. See Chen & Guestrin (2016) for details on
XGBoost, and the vignette for an overview of the package.
ctrdata v0.18: Provides functions for querying, retrieving, and analyzing protocol- and results-related information on clinical trials from two public registers, the European Union Clinical Trials Register and ClinicalTrials.gov. There is a Getting Started Guide and a vignette with examples.
tacmagic v0.2.1: Provides functions to facilitate the analysis of positron emission tomography (PET) time activity curve (TAC) data. See Logan et al. (1996) and Aizenstein et al. (2008) for use cases, and the vignette for a detailed overview.
bulletcp v1.0.0: Provides functions to automatically detect groove locations via a Bayesian changepoint detection method, to be used in the data pre-processing step of forensic bullet matching algorithms. See Stephens (1994) for reference, the vignette for the theory, and Mejia et al. in the most recent issue of Significance for the big picture.
earthtide v0.0.5: Ports the Fortran ETERNA 3.4 program by H.G. Wenzel for calculating synthetic Earth tides using the Hartmann and Wenzel (1994) or Kudryavtsev (2004) tidal catalogs. See the vignette for an introduction.
periscope v0.4.1: Implements an enterprise-targeted, scalable and UI-standardized
shiny framework. There are vignettes for a downloadFile module, downloadablePlot module, downloadableTable module, and the creation of a framework-based application.
reactlog v1.0.0: Provides visual insight into that black box of
shiny reactivity by constructing a directed dependency graph of the application’s reactive state at any point in a reactive recording. See the vignette for an introduction.
shinyhttr v1.0.0: Modifies the
progress() function from the
httr package to let it send output to
progressBar() function from the
CoopGame v0.2.1: Provides a comprehensive set of tools for cooperative game theory with transferable utility, enabling users to create special families of cooperative games, such as bankruptcy games, cost-sharing games, and weighted-voting games. The vignette offers theory and examples.
discfrail v0.1: Provides functions for fitting Cox proportional hazards models for grouped time-to-event data, where the shared group-specific frailties have a discrete non-parametric distribution. See Gasperoni et. al (2018). The vignette shows the math.
fastglm v0.1.1: Provides functions to fit generalized linear models efficiently using
RcppEigen. The iteratively reweighted least squares implementation utilizes the step-halving approach of Marschner (2011). There is a vignette.
hettx v0.1.1: Implements methods developed by Ding, Feller, and Miratrix (2016), and Ding, Feller, and Miratrix (2018) for testing whether there is unexplained variation in treatment effects across observations, and for characterizing the extent of the explained and unexplained variation in treatment effects. There are vignettes on heterogeneous treatment effects and systematic fariation estimation.
mcmcabn v0.1: Implements a structural MCMC sampler for Directed Acyclic Graphs (DAGs). It supports the new edge reversal move from Grzegorczyk and Husmeier (2008) and the Markov blanket resampling from Su and Borsuk (2016), and three priors: a prior controlling for structure complexity from Koivisto and Sood (2004), an uninformative prior, and a user-defined prior. The vignette provides an overview of the package.
networkABC v0.5-3: Implements a new multi-level approximation Bayesian computation (ABC) algorithm to decipher network data and assess the strength of the inferred links between network’s actors. The vignette provides an example.
retrodesign v0.1.0: Provides tools for working with Type S (Sign) and Type M (Magnitude) errors, as proposed in Gelman and Tuerlinckx (2000) and Gelman & Carlin (2014), using the closed forms solutions for the probability of a Type S/M error from Lu, Qiu, and Deng (2018). The vignette shows how to use Type S and M errors in hypothesis testing.
senssobol v0.1.1: Enables users to compute, bootstrap, and plot up to third-order Sobol indices using the estimators by Saltelli et al. (2010) and Jansen (1999), and calculate the approximation error in the computation of Sobol first and total indices using the algorithm of Khorashadi Zadeh et al. (2017). The vignette provides an overview.
RJDemetra v0.1.2: Implements an interface to JDemetra+, the seasonal adjustment software officially recommended to the members of the European Statistical System (ESS) and the European System of Central Banks.
runstats v1.0.1: Provides methods for quickly computing time series sample statistics, including: (1) mean, (2) standard deviation, and (3) variance over a fixed-length window of time-series, (4) correlation, (5) covariance, and (6) Euclidean distance (L2 norm) between short-time pattern and time-series. See the vignette for examples.
lenses v0.0.3: Provides tools for creating and using lenses to simplify data manipulation. Lenses are composable getter/setter pairs for working with data in a purely functional way, which were inspired by the Haskell library
lens ( Kmett (2012) ). For a comprehensive history of lenses, see the
lens wiki and look here for examples.
predict3d v:0.1.0: Provides functions for 2- and 3-dimensional plots for multiple regression models using packages
rgl. It supports linear models (lm), generalized linear models (glm), and local polynomial regression fittings (loess). There is a vignette.