In cardiomoon/moonBook2: Functions and Datasets for the 2nd Book by Keon-Woong Moon

This package contains several functions for interactive plots with ggplot2 and ggiraph package.

Package installation

Package moonBook is avaiable on CRAN and github. Package moonBook2 is available only on github. Please install moonBook2 package using the following R code.

install.packages("devtools")
devtools::install_github("cardiomoon/moonBook")
devtools::install_github("cardiomoon/moonBook2")

ggSpine() : Interactive mosaic plot and spinogram

This function is a extension of graphics::spineplot() function. This function draws a proportional stacked barplot with variable width.

require(moonBook)
require(ggplot2)
require(ggiraph)
require(plyr)
require(reshape2)
require(scales)
require(moonBook2)

ggSpine(data=acs,"Dx","smoking",interactive=TRUE)

Like a bar plot, you can set the parameter position among "stack","dodge" and "fill". You can add labels by setting the parameter addlabel TRUE.

ggSpine(data=acs,"Dx","smoking",addlabel=TRUE,position="stack",interactive=TRUE)
ggSpine(data=acs,"Dx","smoking",addlabel=TRUE,position="dodge",interactive=TRUE)

When used with a continuous variable, it draws a spinogram which is an extension of histogram. In this case, the barwidth is set with a value 1.0(you can adjust the value).

ggSpine(data=acs,Dx,age,addlabel=TRUE,interactive=TRUE)

You can draw a plot with pre-processed data. In this example, we use the rose data in moonBook2 package. You can assign a continuous variable(rose[["value"]]") to y-axis variable and set the parameter stat "identity".

head(rose,10)
ggSpine(rose,"group","Month",yvar="value",stat="identity",width=1,interactive=TRUE)

You can make a "Rose" plot with variable width.

ggSpine(rose,"group","Month",yvar="value",stat="identity",position="stack",width=1,
        polar=TRUE,interactive=TRUE,palette="Reds")

ggRadar() : Interactive Radar chart

Radar chart is a graphical method of displaying three or more continuous data in a 2-dimensional chart. With moonBook::acs data, we can complare all continuous variables in patients with acute coronary syndrome.

ggRadar(acs,groupvar="sex",interactive=TRUE)

Because the distributions of variables are differ, you should rescle the data by set the parameter rescale TRUE.

ggRadar(acs,groupvar="sex",interactive=TRUE,rescale=TRUE)

As you can see, male patients are taller and heavier than female patients. But the bod mass index(BMI) is almost the same.

With mtcars data, you can make a exploratoty plot with this command.

ggRadar(mtcars,groupvar="am",rescale=TRUE,interactive=TRUE)
ggRadar(mtcars,groupvar="cyl",rescale=TRUE,interactive=TRUE)

You can make faceted chart with this code.

mtcars$model=rownames(mtcars)
p<-ggRadar(mtcars[1:9,],rescale=TRUE,groupvar="model",legend.position="none")+
         ylim(0,1)+facet_wrap(~model)
ggiraph(code=print(p))

ggPieDonut() : Combine Pie and Donut plot

You can combine Pie and donut plot easily with ggPieDonut() function. Let's see the browsers data.

str(browsers)

With this data, you can make the PieDonut plot easily.

ggPieDonut(browsers,pies="browser",donuts="version",count="share",interactive=TRUE)

When with categrical variables, ggPieDonut() function calculate the count.

ggPieDonut(acs,"Dx","smoking",interactive=TRUE)

ggPieDonut(mtcars,"cyl","carb",labelposition=0,interactive=TRUE)

ggDount() : Make a Donut plot

ggDonut(browsers,donuts="version",count="share",interactive=TRUE)
ggDonut(browsers,version,count="share",labelposition=0,interactive=TRUE)
ggDonut(acs,Dx,interactive=TRUE)