That's not possible (in DSS 3.x at least). The processors of the prepare recipe are not aware of all the rows (they only work row by row). The Formula is one of these processors. (See more about the prepare recipe in the doc).
As an alternative, you can use a code recipe (SQL, Python, R...) or try to do that with the group recipe.
In a Python notebook, with the library pandas, a df.describe() would do the job.
I hope that helps,
Jeremy, Product Manager at Dataiku