Initial release.
ggstratify() takes a data frame or a matrix that is
already in your session, and nothing else. There is no file-upload
control in the application and a file path is no longer accepted as
dataset. Data read from disk is typed by whatever the
reader guessed, and would then be described on those guesses; read the
file yourself, look at the result, and pass the object.
The name the data was passed under is what the generated code
refers to it by, so ggstratify(cohort) prints code that
starts d <- cohort.
Previews are subsampled only above one million rows for scatter and line plots and ten million for the plot types that summarise their rows first, and never for Kaplan-Meier curves.