Skip to content
Automatic plotting with ggplot
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
R
inst/tests
man-roxygen
man
scratch
tests
.Rbuildignore
.gitignore
DESCRIPTION
NAMESPACE
README.md
autoplot.Rproj

README.md

autoplot

ggplot2 is a versatile plotting package which allows to produce almost any kind of plot from data stored as a data.frame, by combining unit elements. However, it requires the user to design the plot entirely, from scratch. Many R functions for statistical analyses (linear models, mutlivariate analyses, etc.) output objects of a given class and allow to easily plot classic diagnostics using plot(), by defining a specialized method for this generic function. This package aims at reproducing this functionality in ggplot2 while benefiting from its improved versatility. It provides two sets of methods : (i) the augment() methods extract data from the original object and format it as a data.frame (see package broom); (ii) the autoplot() methods use these data.frames and leverage ggplot2 to produce diagnostic plots.

Installation

The package is under development and is not on CRAN yet. To install it, the simplest method is therefore

# install.packages("devtools")
devtools::install_github("jiho/autoplot")

Development

All suggestions and improvements are welcome. Suggestions are preferred under the form of Github "issues" (see the tab above, on the top right). Improvements should be submitted under the form of "pull-requests" (again, see the tab above).

The best way to create a pull-request (according to me) is:

  1. fork the repository on GitHub's website. It creates your very own copy of autoplot

  2. clone my version the project

     git clone https://github.com/jiho/autoplot.git
    
  3. create a branch

     git branch myfeature
     git checkout myfeature
    
  4. make the changes you want and commit them to the branch

     <code, code>
     git commit -a
    
  5. push the branch to your github repository

     git remote add mine https://github.com/<yourname>/autoplot.git
     git push myfeature mine
    
  6. visit your branch on GitHub's website and create a pull request. It should automatically be made against my master branch.

This way, you can easily update the master branch to get my latest changes and base your next contribution on it (which is necessary for things not to break):

# update the master branch to match mine
git checkout master
git pull

# create a new branch
git branch myOtherFeature
git checkout myOtherFeature

# create a new change
<code, code>
git commit

# push it online
git push myOtherFeature mine

Technically, the documentation is written with roxygen, unit tests are performed with testthat. You should use that. It's too cool to miss out.

You can’t perform that action at this time.