Count should always return a data-frame

[holgerbrandl]

Currently df.rowsCount() and df.count() do the same ie are redudnant for ungrouped data-frames. However, for sake of consistency, count() should always return a data-frame irrespective if data is grouped or not.

In particular, this type of consistency made dplyr hugely popular, because it guaranteed that all major functions work df-input -> df-output without exception (grouping or not)

[koperagen]

Hm, but how do they summarise grouped dataframes?
Right now, it works like this:

groupBy.aggregate {
    // this = dataframe
    count() into "count"
}

If count() returns a dataframe, we need to get actual value from it or introduce overloads, maybe

[holgerbrandl]

It does not work that way count is just a shortcut for summarize (the dplyr equivalent for aggregate) something along count = fun(tbl){ summarize(tbl, n=n())}. To get the number of rows in an aggregate statement the do not use count but a special context function (which is only available within summarize called n() which returns the number of rows per group (or in total for ungrouped data.

By doing so, they provide a very consistent overall API experience irrespective if data is grouped or not.