qwraps2

A collection of helpful functions for summarizing data and formatting results. These or similar functions can be found in other R packages on github.com or on CRAN. However, this is a collection of methods I have generated to help with particular projects I have worked on over the last several years. Using my original qwraps package as the foundation, this work is aimed at simplicity and ease of use.

Some of the tools provided by qwraps2 are:

• Formatting results:

  • frmt, frmtci, and frmtp make it easier to consistently format numeric values, confidence intervals, and p-values in reports.
  • qable is a wrapper around knitr::kable with my preferred defaults.
  • summary_table is used to generate data summary tables in markdown or LaTeX with the look of well formatted tables from the xtables or Hmisc::latex.
  • mean_ci, mean_sd, median_iqr, n_perc make it easy to report formatted summary statistics.

• Plotting: Tools for building specific types of plots in ggplot2

  • qacf: Autocorrelation plots
  • qblandaltman: Bland Altman plots
  • qkmplot: Kaplan-Meier plots
  • qroc: Receiver Operator Curves
  • ggplot2_extract_legend: extract the legend from a ggplot.

• R Programming and Development

  • lazyload_cache Load the cache generated by knitr into an interactive session.

• Other Tools:

  • confusion_matrix: building and generate statistics for confusion matrices.
  • logit and invlogit: quick logit and inverse logit functions
  • ll: a variant for ls()

Contribute!

If you have a particular task or function used for data summaries or for reporting that you think would be helpful to include in this package please fork the repo, add the feature, and send me a pull request.

Building the package

Use the makefile. RStudio users, you will find a qwraps2.Rproj that will set the default to use the makefile in the build. My preferred IDE is neovim with the Nvim-R plug-in and I prefer to work on a Debian system. After cloning the repo, a simple call

make

will build the package. This includes generating man files via roxygen2, building the vignettes, and then building the package via R CMD build .. Passing arguments to R CMD build can be done too. For example, building the package without the vignettes is done via:

make build-options=--no-build-vignettes

Install

From CRAN

Download and install from The Comprehensive R Archive Network (CRAN).

install.packages("qwraps2", repo = "http://cran.rstudio.com")

Developmental

Install the development version of qwraps2 directly from github via the remotes package:

if (!("remotes" %in% rownames(installed.packages()))) {
  warning("installing remotes from https://cran.rstudio.com")
  install.packages("remotes", repo = "https://cran.rstudio.com")
}

remotes::install_github("dewittpe/qwraps2", build_vignettes = TRUE)

NOTE: If you are working on a Windows machine you will need to download and install Rtools.

Cloned repo

Install with GNU make

make install