One hundred forty-eight new packages made it to CRAN in April. Here are my “Top 40” picks in nine categories: Computational Methods, Data, Machine Learning, Medicine, Science, Statistics, Time Series, Utilities, and Visualization.
Computational Methods
JuliaConnectoR v0.6.0: Allows users to import Julia packages and functions in such a way that they can be called directly as as R functions.
RcppBigIntAlgos: v0.2.2: Implements the multiple polynomial quadratic sieve (MPQS) algorithm for factoring large integers and a vectorized factoring function that returns the complete factorization of an integer. See Pomerance (1984) and Silverman (1987) for background and this Microsoft post for an explanation.
smoothedLasso v1.0: Implements the smoothed LASSO regression using the method of Nesterov (2005).
Data
daqape v0.3.0: Provides a variety of methods to identify data quality issues in process-oriented data. There is an Introduction.
DSOpal v1.1.0: is the DataShield implementation of Opal, the data integration application for biobanks by OBiBa, open source software for epidemiology.
epuR v0.1: Provides functions to collect data from the the economic policy uncertainty website. See the vignette.
hystReet v0.0.1: Implements an API wrapper for the Hystreet project which provides pedestrian counts for various cities in Germany. See the vignette to get started.
rGEDI v0.1.7: Provides a set of tools for downloading, reading, visualizing and processing GEDI Level1B, Level2A and Level2B data. see the vignette to get started.
Machine Learning
catsim v0.2.1: Computes structural similarity metrics for binary and categorical 2D and 3D images including Cohen’s kappa, Rand index, adjusted Rand index, Jaccard index, Dice index, normalized mutual information, or adjusted mutual information. See Thompson & Maitra (2020) for background and the vignette for an introduction.
klic v1.0.2: Implements a kernel learning integrative clustering algorithm which allows combining multiple kernels, each representing a different measure of the similarity between a set of observations. There is an Introduction.
MIDASwrappeR V0.5.1: Provides a wrapper for the C++ implementation of the MIDAS algorithm described in Bhatia et al. (2020) for graph like data. See the Introduction.
VUROCS v1.0: Calculates the volume under the ROC surface and its (co)variance for ordered multi-class ROC analysis as well as certain bivariate ordinal measures of association.
WeightSVM v1.7-4: Provides functions for subject/instance weighted support vector machines (SVM). It uses a modified version of libsvm and is compatible with e1071 package. Look here for some background.
Medicine
covid19.analytics v1.1: Provides functions to load and analyze COVID-19 data from the Johns Hopkins University CSSE data repository. It includes functions to visualize cases for specific geographical locations, generate interactive visualizations and produce a SIR model. See the vignette for an introduction.
covid19france Provides functions to import, clean and update French COVID-19 data from opencovid19-fr.
interactionR v0.1.1: Produces a publication-ready table that includes all effect estimates necessary for full reporting effect modification and interaction analysis as recommended by Knol & Vanderweele (2012), estimates confidence interval additive interaction measures using the delta method Hosmer & Lemeshow (1992), the variance recovery method Zou (2008), or percentile bootstrapping Assmann et al. (1996).
RCT v1.0.2: Provides tools to facilitate the process of designing and evaluating randomized control trials, including methods to handle misfits, power calculations, balance regressions, and more. For background see Athey et al. (2017). The vignette describes how to use the package.
Science
rasterdiv: Provides functions to calculate indices of diversity on numerical matrices based on information theory. The rationale behind the package is described in Rocchini et al. (2017). See the vignette for an extended example.
SSHAARP v1.0.0: Processes amino acid alignments from the IPD-IMGT/HLA database to identify user-defined amino acid residue motifs shared across HLA alleles, calculate the frequencies of those motifs, and generate global frequency heat maps that illustrate the distribution of each user-defined map around the globe. See the vignette for an introduction.
Statistics
BayesSampling v1.0.0: Provides functions for applying the Bayes Linear approach to finite populations with the simple random sampling, stratified simple random sampling designs, and to the ratio estimator. See Gonçalves et al. (2014) for background and the vignettes: BLE_Ratio, BLE_Reg, BLE_SRS, BLE_SSRS, and BayesSampling.
cort v0.3.1: Provides S4 classes and methods to fit several copula models including empirical checkerboard copula Cuberos et. al (2019) and the Copula Recursive Tree algorithm proposed by Laverny et. al (2020). There are vignettes on the Empirical Checkerboard Copula, the Copula Recursive Tree, the Empirical Checkerboard Copula with known margins, and the convex mixture of m-randomized checkerboards.
ExpertChoice v0.2.0: Implements tools for designing efficient discrete choice experiments. See Street et. al (2005) for some background. There is an Practical Introduction and a vignette with some theory.
genscore v1.0.2: Implements the generalized score matching estimator from Yu et al. (2019) for non-negative graphical models with truncated distributions, and the estimator of Lin et al. (2016) for untruncated Gaussian graphical models. See the vignette.
hmma v1.0.0: Provides functions to fit Bayesian asymmetric hidden Markov models. HMM-As are similar to regular HMMs, See Bueno et al. (2017) for background and the vignette for and introduction.
lmeInfo v0.1.1: Provides analytic derivatives and information matrices for fitted linear mixed effects models and generalized least squares models estimated using lme() and gls() as well as functions for estimating the sampling variance-covariance of variance component parameters and standardized mean difference effect sizes. See Pustejovsky et al. (2014) and the vignette.
metapower v0.1.0: Implements a tool for computing meta-analytic statistical power for main effects, tests of homogeneity, and categorical moderator models. Have a look at Pigott (2012), Hedges & Pigott (2004), or Borenstein et al. (2009) for background and the vignett to get started.
sasLM v0.1.3: Implements the SAS procedures for linear models: GLM, REG, ANOVA. The sasLM functions produce the same results as the corresponding SAS procedures for nested and complex models.
sdglinkage 0.1.0: Provides a tool for synthetic data generation that can be used for linkage method development. There is an Overview and vignettes on Real and Synthetic Identifiers, Gold Standard File and Linkage Files, Synthetic Data Generation and Evaluation.
starm v0.1.0: Estimates the coefficients of the two-time centered autologistic regression model described in Gegout-Petit et al. (2019). The vignette describes the theory.
Time Series
ConsReg v0.1.0: Provides functions to fit regression and generalized linear models with autoregressive moving-average (ARMA) errors for time series data. There is a vignette.
simITS v0.1.1: Implements the method of Miratrix (2020) to create prediction intervals for post-policy outcomes in interrupted time series. It provides methods to fit ITS models with lagged outcomes and variables to account for temporal dependencies and then to simulate a set of plausible counterfactual post-policy series to compare to the observed post-policy series. See the vignette.
Utilities
dreamerr v1.1.0: Implements tools to facilitate package development by providing a flexible way to check the arguments passed to functions. See the vignette for details.
flair v0.0.2: Facilitates formatting and highlighting of R source code in a R Markdown based presentation. The vignette shows how.
J4R v1.0.7: Makes it possible to create Java objects and to execute Java methods from the R environment. The JVM is handled by a gateway server which relies on the Java library j4r.jar.
waldo v0.1.0: Provides functions to compare complex R objects and reveal the key differences. It was designed primarily for use in testing packages.
Visualization
anglr v0.6.0: Extends rgl conversion and visualization functions to mesh3d to give direct access to generic 3D tools and provide a full suite of mesh-creation and 3D plotting functions. See the vignette
brickr v0.3.4: Uses tidyverse functions to generate digital LEGO models and convert image files into 2D and 3D LEGO mosaics. There are vignettes for building mosaics and for generating models from mosaics, programs, tables, and by piece type.
survCurve v1.0: Provides functions to enhance plots created with the survival and mstate packages. See the vignette for examples.
textplot v0.1.2: Provides functions to visualize complex relations in texts by displaying text co-occurrence networks, text correlation networks, dependency relationships and text clustering. The vignette provides examples.
You may leave a comment below or discuss the post in the forum community.rstudio.com.