by Joseph Rickert

September was a bit of a slow month for new R packages. Only 96 new packages showed up on CRAN. Nevertheless, I have picked out 23 for special mention which I have listed in 5 categories. I used the same selection criteria as I described in the post for August picks.

## Data and Interfaces

- darksky v1.0.0: Provides an interface to the Dark Sky API which allows you to look up weather anywhere on the globe.
- etseed v0.1.0: Provides a client to interface to the etcd key value store, a database written in Go.
- LendingClub v0.1.2: Lets investors manage their LendingClub investments from R.
- sparklyr V0.4: Allows R users to connect, provision and interface to Apache Spark. Detailed examples using MLlib and H2O are available on the RStudio site.
- trelloR V0.1.0: Provides access to the Trello API. The vignette explains how to retrieve data from public and private Trello boards.
- XRPython V0.7: A Python interface structured according to the general form of the package XR described in John Chamber’s book Extending R.

## Machine Learning

- ensembleR v0.1.0: Facilitates constructing ensemble models from machine learning models available in the caret package. There is a vignette to get started.
- exprso v0.1.7: Provides a framework for supervised machine learning customized for biologists. There are several vignettes including a cheatsheet.

- lowmemtkmeans v0.1.0: Implements trimmed k-means clustering with low memory use.
- Textmining V0.0.2: Provides functions for text and topic mining. Full functionality requires installing TreeTagger.

## Plots and Visualizations

- plotluck v1.0.1: Is an intelligent tool built on top of ggplot2 that automatically generates plots from dataframes based on users providing variables to plot.
- plotwidgets V0.4: Provides functions to produce small, self contained plots for use in larger plots.

```
library(plotwidgets)
plot.new()
par(usr = c(-1, 1, -1, 1))
hues <- seq(0, 360, by = 30)
pos <- a2xy(hues, r = 0.75)
for (i in 1:length(hues)) {
cols <- modhueCol(pal, by = hues[i])
wgPlanets(
x = pos$x[i],
y = pos$y[i],
w = 0.5,
h = 0.5,
v = v,
col = cols
)
}
pos <- a2xy(hues[-1], r = 0.4)
text(pos$x, pos$y, hues[-1])
```

## Statistics

- Barycenter v1.0.0: Provides algorithms to compute the Wasserstein barycenter, the mean of a set of empirical probability measures.
- musica v0.1.3: Provides functions for working with multivariate time series and custom time scales. There is a vignette to help you get started.
- nhstplot v1.0.0: Provides functions to graphically illustrate the most common null hypothesis significance tests. The vignette provides some examples, e.g.:

```
library(nhstplot)
plotftest(4,3,5)
```

- nimble v0.6-1: Allows R programmers to write statistical models in the BUGS language. NIMBLE is built in R but compiles in C++. There is extensive documentation at http://www.nimble.org
- Rdice v1.0.1: Allows conducting sophisticated dice rolling and coin tossing experiments including experiments with Efron like Nontransitive dice. Have a look at the vignette.
- splines2 V0.1.0: Provides functions for constructing a variety of splines that are not available in the splines package including B-splines, M-splines, I-splines, C-splines, and the integral of B-splines. There is a vignette.
- scanstatistics v0.1.0: Provides scan statistics functions to detect anomalous clusters in spatial or space-time data. The vignette describes the methodology and presents examples as well.
- thief v0.2: Implements methods for generating forecasts at different temporal frequencies using hierarchical time series.

## Misc

- GeneralTree v0.0.1: Implements a general tree data structure in R. There is a tutorial.
- radiant V0.6.0: Radiant is a platform for business analytics based on R and Shiny.
- RDocumentation v0.6: A wrapper for R’s documentation so that help documentation will appear as it does on http://www.rdocumentation.org.
- tableMatrix v0.8: Provides 2 classes extending data.table: tableList and tableMatrix.

With this post, I am up to date with new CRAN packages. I hope to make my package picks a regular, monthly feature of this blog.

You may leave a comment below or discuss the post in the forum community.rstudio.com.