R package ggplotAssist

Keon-Woong Moon

2017-10-24

The 'ggplotAssist' is an RStudio addin for teaching and learning plot generation using the 'ggplot2' package. You can learn each steps of plot generation - aesthetics mapping, select geometries, add scales, apply theme - by clicking your mouse without coding. You can see the resultant plot and see the each steps of plot layer by layer. You get resultant code for ggplot. 

Prerequisite

You have to install the developmental version of R package editData from github.

#install.packages("devtools")
devtools::install_github("cardiomoon/editData")

Install package

You can install ggplotAssist package from github.

#install.packages("devtools")
devtools::install_github("cardiomoon/ggplotAssist")

Usage: As an RStudio Add-in

This addin can be used to interactively generate a ggplot using ggplot2 package. The intended way to use this is as follows:

  1. Highlight a symbol naming a data.frame or a tibble in your R session, e.g. msleep(1). Execute this addin(arrow), to interactively manipulate it.

plot of chunk unnamed-chunk-3

  1. You can see a brower window. You can see the data name(1) and R code for ggplot(2). Select x(3) and bodywt(4) to map bodywt as a x-axis variable.

plot of chunk unnamed-chunk-4

  1. You can see the R code for ggplot(1). Select y and brainwt to map the y-axis variable.

plot of chunk unnamed-chunk-5

  1. To add geoms to ggplot, press geom button(1) and select geom_point(2). You can mapping or setting the aesthetics of geom_point. You can see the R code for this layer(3). In the lower part of window, you can see two R codes and two plots. In the lower left portion, you can see the R code for plot(4) and resultant plot(5). In the lower right portion, you can see the R code for plot with R code for layer under construction(6) and resultant plot(7). This is a plot preview.

plot of chunk unnamed-chunk-6

  1. If you finish to make a layer, press Add Layer button(1) to add the layer. You can see added layers(2). You can delete a layer with delete layer button after select a layer to delete.

plot of chunk unnamed-chunk-7

  1. To add theme, press theme button(1) and select theme_bw(2). You can see the code for theme_bw(3) and plot preview(4,5). Press Add Layer button(6) to add the code. plot of chunk unnamed-chunk-8

  2. To apply a log scale to x-axis, press scale button(1) and select scale_x_log10 function(2). You can set the arguments of scale_x_log10() function(3). You can see the code for scale(4) and plot preview(5,6). To add math format to the log scale, press Add math-format button(9). You can repeat this step to apply a log scale to y-axis. plot of chunk unnamed-chunk-9

  3. You can add logticks to your plot. Press annotate button(1) and select annotation_logticks function(2). You can see the arguments and default values of this function(3). Set the side argument trbl(4). You can see the R code for this layer(5) and plot preview(6.7). Press Add Layer button(8) to add this layer.

plot of chunk unnamed-chunk-10

  1. You can hide the minor grid lines because they don’t align with the ticks. Press theme button(1) and select the theme function(2). The theme() function has a lot of arguments. Select panel.grid.minor argument(3) and select element_blank()(4) to hide the minor grid line.

plot of chunk unnamed-chunk-11

  1. You can see this plot layer by layer. Press the Layer by layer button(1) and use the sliderInput to see the plot layer by layer. You can animate this plot by click the arrowhead(2). If you want to get this R code for plot, press Done button(3).

plot of chunk unnamed-chunk-12

  1. When you’re done, the code for the ggplot will be emitted at the cursor position(scarlet rectangle).

plot of chunk unnamed-chunk-13

Usage: As a regular function

You can use the ggplotAssist() function as a regular function, e.g. in a command line.

result <- ggplotAssist(mtcars)

Other topics

How to use guide_colorbar()?

http://rpubs.com/cardiomoon/322415

Recursive shiny module for functionals

There are many functions that takes a function as an input : Functionals. To handle a functional in a shiny app, you have to make a shiny module that allows recursive call. I have included an recursive shiny module textFunction in my package ggplotAssist. The UI of textFunction shiny module is textFunctionInput and the server function is textFunction. I also included two toy shiny app to demonstrate the recursive shiny module.

You can read the full story at https://github.com/cardiomoon/ggplotAssist/blob/master/textFunctionInput.md