kylelundstedt/rpackages

This package attaches the recommended packages for use on the IndustryVault platform

IndustryVault's Recommended R Packages

This is an R package that just attaches the IndustryVault-recommended packages (industryvault-packages) for use on our platform. It includes most of Hadley Wickham's packages (colloquially known as the "Hadleyverse"). It's based on a similar package by Tony Boyles.

Installation

install.packages("devtools")
library("devtools")
install_github("IndustryVault/Rpackages")

Use

library("Rpackages") 
# All of the IndustryVault-recommended R packages are now available in your environment.
# There's no need to call library("dplyr"), etc.

What Happens

When you install an R Package, R checks the DESCRIPTION file for dependencies. If you have unmet dependencies, R tries to install them from CRAN. Then, whenever you load the package, R makes those dependencies available. This package just "depends on" all of the packages that the IndustryVault team recommends, despite the fact that it doesn't do anything.

Recommended Packages

The three most important R packages used on the IndustryVault platform are dplyr, ggplot2, and rmarkdown. For R users new to our platform, these are the first packages with which you should become familiar.

That said, here is the complete list of packages in the Rpackages package installs and loads. Note that packages associated with the "Hadleyverse" are denoted with (H), and packages hosted on GitHub.com (not CRAN) are denoted with (G). The packages installed include:

Ingest

• readr: read flat files (csv, tsv, fwf) into R (H)
• readxl: read excel files (.xls and .xlsx) into R (H)
• haven: read SPSS, Stata and SAS files from R (H)
• rvest: simple web scraping for R (H)
• httr: download web data ala the curl command, customised to modern web APIs (H)
• wakefield: generate random data sets

Manipulation

• dplyr: data manipulation ala plyr, specialised for data frames (H)
• multidplyr: a backend for dplyr that partitions a data frame across multiple cores (G, H)
• testdat: a package to run unit tests on tabular data (G)
• assertr: a suite of functions designed to verify assumptions about data
• tidyr: easily tidy data with spread and gather functions (H)
• lubridate: make working with dates in R just that little bit easier (H)
• stringr: wrapper for R string functions to make them more consistent, simpler and easier to use (H)
• broom: convert statistical analysis objects from R into tidy format
• RPostgreSQL: R interface to the PostgreSQL database, inc. AWS Redshift (rstats-db/RPostgres in development)
• chunked: chunkwise text-file processing for 'dplyr'
• ezsummary: pre-programming the common dataset summary task

Visualization

• ggplot2: powerful plotting library in R based on the Grammar of Graphics (H)
• ggalt: extra coordinate systems, geoms, and statistical transformations for ggplot2 (H)
• ggsubplot: embed subplots in ggplot2 graphics in R (H)
• dotwhisker: generate dot-and-whisker plots of regression results
• ggmap: plotting maps in R with ggplot2 (H)
• ggcounty: generate ggplot2 geom_map county maps
• choroplethr: simplifies the creation of choropleths (thematic maps) in R
• choroplethrZip: simplifies data and visualization of US ZIP Codes (G)
• statebins: "map-like" alternative to choropleths of US States
• gtable: tools to make it easier to work with “tables” of graphics objects (grobs) (H)
• ggthemes: ggplot themes and scales
• scales: more graphical scales, for use on chart axes, etc (H)
• directlabels: direct labels for multicolor plots in lattice or ggplot2
• ggrepel: repel overlapping text labels away from each other (G)
• ggvis: web-based data visualizaton ala ggplot2
• svglite: lightweight svg graphics (web optimal) for R (H)
• bigvis: exploratory data analysis for large datasets (10-100 million observations) (G, H)
• metricsgraphics: htmlwidget interface to the MetricsGraphics.js D3 chart library
• leaflet: JavaScript library for mobile-friendly interactive maps based on htmlwidgets (G)
• rafalib: interesting collection of functions for data exploration

Reporting

• rmarkdown: dynamic documents for R (H)
• knitr: a general-purpose tool for dynamic report generation in R (H)
• knitcitations: generate citations for knitr markdown and html files
• captioner: generate figure/table numbers and captions for RMarkdown docs
• DiagrammeR: create graph diagrams and flowcharts using R
• pixiedust: broom-compatiable table formatting
• stargazer: well-formatted regression and summary statistics tables
• xtable: export Tables to LaTeX or HTML (alternative to stargazer)
• tufterhandout: output formats for Tufte-style handouts in pdf and html for Rmarkdown
• revealjs: RMarkdown custom format for reveal.js HTML presentations (G)
• bookdown: package and set of conventions for making books with Rmarkdown and RStudio (G)

Programming

• devtools: tools to make an R developer's life easier (H)
• magrittr: R package to bring forward-piping features ala F#’s |> operator.
• testthat: an R package to make testing fun (H)
• rex: friendly regular expressions for R (G)
• htmlwidgets: HTML Widgets for R, easily creating R bindings to JavaScript libraries
• assertthat: user-friendly assertions for R (H)
• lazyeval: a strategy for doing non-standard evaluation (NSE) in R (H)
• profvis: visualise line profiling results in R (H)
• lintr: static code analyis (i.e., lint) for R
• foreach: looping construct (i.e., for loop) that support parallel execution
• doParallel: foreach parallel adaptor for the 'parallel' package
• snippr: manage, share, and install RStudio code snippets (G)
• slackr: send messages to Slack channels/users from R
• rpushbullet: send messages to your computer, phone, or tablet using Pushbullet service

TODO

• decide on implementing this package using drat, packrat, or some other method. (Not checkpoint due to GitHub repos)

• add vignette for use of three most important packages dplyr, ggplot2, and rmarkdown

• add vignette for examples of all packages

• best practices ?

• consider adding separate packages for datasets; see, for example:

  • noncensus: self-contained U.S. Census Region and Demographic Data
  • rUnemploymentData: data and functions for dealing with US Unemployment Data in R
  • tigris: download and use Census TIGER/Line shapefiles in R

• safety testing

  • find out if it's safe to add rdrop2: Dropbox Interface from R
  • should we allow user-installed packages on AWS?
  • can we prevent data export on AWS from R and RStudio?

• GitHub repo installation code is below:

  install_github( c("arilamstein/choroplethrZip", "dgrtwo/snippr", "eddelbuettel/rpushbullet", "edwindj/chunked", "hadley/bigvis", "hadley/svglite", "haozhu233/ezsummary", "hrbrmstr/ggcounty", "jjallaire/revealjs", "karthik/rdrop2", "kevinushey/rex", "nutterb/pixiedust", "ramnathv/slidify", "ramnathv/slidifyLibraries", "ropensci/testdat", "rstudio/bookdown", "trinker/wakefield", "walkerke/tigris") )