September was another relatively slow month for new package activity on CRAN: “only” 126 new packages by my count. My Top 40 list is heavy on what I characterize as “utilities”: packages that either extend R in some fashion or make it easier to do things in R. This month, the packages I selected fall into eight categories: Data, Finance, Machine Learning, Science, Statistics, Time Series, Utilities and Visualization.
crfsuite v0.1.1: Wraps the CRFsuite library allowing users to fit a conditional random field model. The focus is Natural Language Processing, and there are models for named entity recognition, text chunking, part of speech tagging, intent recognition, and classification. The vignette shows how to use the package.
ELMSO v1.0.0: Implements the algorithm described in Paulson, Luo, and James (2018); see here for a full-text version of the paper. The algorithm allocates budget across a set of online advertising opportunities.
newsmap v0.6: Implements a semi-supervised model for geographical document classification ([Watanabe (2018)])(doi:10.1080⁄21670811.2017.1293487) with seed dictionaries in English, German, Spanish, Japanese, and Russian. See the README for an example.
splinetree v0.1.0: Provides functions to build regression trees and random forests for longitudinal or functional data using a spline projection method. Implements and extends the work of Yu and Lambert (1999). There is an Introduction and vignettes on trees and forests.
stylest v0.1.0: Provides functions to estimate the distinctiveness in speakers’ (authors’) style. Fits models that can be used for predicting speakers of new texts. See Spirling et al (2018) for the details and the vignette for an example on how to use the package.
conStruct v1.0.0: Provides a method for modeling genetic data as a combination of discrete layers, within each of which relatedness may decay continuously with geographic distance. There are vignettes for formatting data, model construction, and on running and visualizing
episcan v0.0.1: Provides some efficient mechanisms to scan epistasis in genome-wide interaction studies (GWIS), and supports both case-control status (binary outcome) and quantitative phenotype (continuous outcome) studies. See Kam-Thong and Cxamara et al. (2011), Kam-Thong and Pütz et al. (2011), and the vignette.
ahpsurvey v0.2.2: Implements the Analytic Hierarchy Process, a versatile multi-criteria decision-making tool introduced by Saaty (1987) that allows decision-makers to weigh attributes and evaluate alternatives presented to them. The vignette provides examples.
empirical v0.1.0: Implements empirical univariate probability density functions (continuous functions) and empirical cumulative distribution functions (step functions or continuous). The vignette provides examples.
MetaStan v0.0.1: Provides functions to perform Bayesian meta-analysis using
Stan. Includes binomial-normal hierarchical models and option to use weakly informative priors for the heterogeneity parameter and the treatment effect parameter, which are described in Guenhan, Roever, and Friede (2018). The vignette contains an example.
Opt4PL v0.1.1: Provides functions to obtain and evaluate various optimal designs for the 3-, 4-, and 5-parameter logistic models. The optimal designs are obtained based on the numerical algorithm in Hyun, Wong, Yang (2018).
rwavelet v0.1.0: Provides functions to perform wavelet analysis (orthogonal and translation invariant transforms) with applications to data compression or denoising. Most of the code is a port of the
MATLAB Wavelab toolbox written by Donoho, Maleki and Shahram. The vignette provides examples.
samplingBigData v1.0.0: Provides methods for sampling large data sets, including spatially balanced sampling in multi-dimensional spaces with any prescribed inclusion probabilities. Written in C, it uses efficient data structures such as k-d trees that scale to several million rows on a modern desktop computer.
survivalAnalysis v0.1.0: Implements a high-level interface to perform survival analysis, including Kaplan-Meier analysis and log-rank tests and Cox regression. There are vignettes for univariate and multivariate survival analyses.
ungroup v1.1.0: Provides functions to implement a penalized composite link model for efficient estimation of smooth distributions from coarsely binned data. For a detailed description of the method and applications, see Rizzi et al. (2015). The vignette provides examples.
bayesdfa v0.1.0: Implements Bayesian dynamic factor analysis, a dimension-reduction tool for multivariate time series, with
Stan. The vignette shows how to identify extremes and latent regimes with
atable v0.1.0: Provides functions to create tables for reporting clinical trials, calculate descriptive statistics and hypotheses tests, and arrange the results in a table with
Word. The vignette provides examples.
broom.mixed v0.2.2: Converts fitted objects from various R mixed-model packages into tidy data frames along the lines of the
piggyback v0.0.8: Works around git’s 50MB commit limit to allow larger (up to 2 GB) data files to piggyback on a repository as assets attached to individual GitHub releases. There is a package overview and a vignette on alternatives.
pysd2r v0.1.0: Uses
reticulate to implement an interface to the
pysd toolset, provides a number of
pysd functions, and can read files in
xmile formats. The vignette provides an overview.
radix v0.5: Provides functions to format scientific and technical articles for the web with Radix reader-friendly typography, flexible layout options for visualizations, and full support for footnotes and citations.
salty v0.1.0: Lets users take real or simulated data and salt it with errors commonly found in the wild, such as pseudo-OCR errors, Unicode problems, numeric fields with nonsensical punctuation, bad dates, etc. See README for examples.
customLayout v0.2.0: Offers an extended version of the
graphics::layout() function that also supports
grid graphics, allowing users to create complicated drawing areas for multiple elements by combining much simpler layouts. The vignette for
ggparliament v2.0.0: Provides parliament plots to visualize election results as points in the architectural layout of the legislative chamber. There are vignettes for arranging parliament, basic plots, drawing majorities, emphasizing parliamentarians, faceting,
hanging seats, highlighingt government, and labeling parties.