The goal of ggls is to add letter indicating the significance of differences of results among treatments displayed in ggplot barplots and boxplots. It doesn’t do that part yet, but it does offer a comparison maker to plug into the ggpubr package as well as a testing assumptions function which tests common assumptions about datasets required by certain statistical tests.
Installation
You can install the development version of ggls from GitHub with:
# install.packages("devtools")
devtools::install_github("beabock/ggls")Example
If you want to use the ggpubr package, you may need to create a list of comparisons for the groups you want to compare. If you want to compare all groups to each other, this command does that.
library(ggls)
comparison_maker(ToothGrowth, supp)
#> [[1]]
#> [1] "OJ" "VC"Additionally, you may want to test the common assumptions of common statistical tests on your dataset to know if it’s okay to use those tests on your dataset. The test_assumptions function does this.
test_assumptions(ToothGrowth, supp, len)
#> [1] "Shapiro Test indicates non-normal distribution in the data"
#> [1] "Variance Test indicates equal variances in the data"
#> Assumption Decision
#> [1,] "Shapiro Normality" "Fail"
#> [2,] "Levene Variance" "Pass"You’ll still need to render README.Rmd regularly, to keep README.md up-to-date. devtools::build_readme() is handy for this.