Open IndrajeetPatil opened 2 years ago
My preference would be to implement a set of descriptive functions with general names (e.g., distribution_mean
, distribution_sd
, distribution_mad
, distribution_cv
) that consistently support a common set of useful arguments:
weights
trim
method
(e.g., to compute maximum likelihood, unbiased, or robust versions of a stat).and perhaps other arguments.
Small point: I really hate how "standard error" is commonly used to refer specifically to the SE of the mean, so if we added something like that, I would prefer it be something like distribution_sem()
or distribution_se_mean()
.
Where are we with this? Currently, the "Data properties" section contains these functions:
It would be useful to have a clear list of functions to add or import from other packages.
Maybe also some of the summary statistics from sjstats can be implemented in datawizard?
Originally posted by @strengejacke in https://github.com/easystats/datawizard/issues/12#issuecomment-854410266