ggmosaic was designed to create visualizations of categorical data and is capable of producing bar charts, stacked bar charts, mosaic plots, and double decker plots.
You can install ggmosaic from github with:
# install.packages("devtools")
devtools::install_github("haleyjeppson/ggmosaic")
library(ggmosaic)
library(dplyr)
happy %>%
mutate(finrela = forcats::fct_recode(finrela,
"far below " = "far below average",
" below" = "below average",
"average" = "average",
"above " = "above average",
"l\n far above" = "far above average")) %>%
ggplot() +
geom_mosaic(aes(x = product(finrela), fill=health), show.legend = FALSE) +
theme_mosaic() +
scale_fill_manual(values = c("#4575B4", "#ABD9E9", "#FEE090", "#F46D43"))
In geom_mosaic()
, the following aesthetics can be specified:
weight
: select a weighting variable
x
: select variables to add to formula
x = product(var2, var1, ...)
alpha
: add an alpha transparency to the selected variable
x
, it will be added to the formula in the first positionfill
: select a variable to be filled
x
, it will be added to the formula in the first position after the optional alpha
variable.conds
: select a variable to condition on
conds = product(cond1, cond2, ...)
These values are then sent through repurposed productplots
functions to create the desired formula: weight ~ alpha + fill + x | conds
.
Since the initial release of ggmosaic, ggplot2 has evolved considerably. And as ggplot2 continues to evolve, ggmosaic must continue to evolve alongside it. Although these changes affect the underlying code and not the general usage of ggmosaic, the general user may need to be aware of compatibility issues that can arise between versions. The table below summarizes the compatibility between versions.
ggmosaic | ggplot2 | Axis labels | Tick marks |
---|---|---|---|
0.3.3 | 3.3.3 | x | x |
0.3.0 | 3.3.0 | x | x |
0.2.2 | 3.3.0 | Default labels are okay, but must use scale_*_productlist() to modify |
No tick marks |
0.2.2 | 3.2.0 | Default labels okay, but must use scale_*_productlist() to modify |
x |
0.2.0 | 3.2.0 | Default labels are wrong, but can use labs() to modify |
x |