so
.N()
.SD()
.I()
etc., will all have exactly the same meaning as their symbol forms
That would address https://github.com/Rdatatable/data.table/issues/4318
I feel like this would cause confusion by users. The confusion being: why do some examples have .N() while other examples have .N.
And while .N is one thing, there are a lot of use cases of subsetting with .SD and .I. Do we want to support both .SD()[var1 > 3, sum(var2)] in addition to .SD[var1 > 3, sum(var2)]?
We don't need to put those into examples, we just need to translate them internally.
Most helpful comment
I feel like this would cause confusion by users. The confusion being: why do some examples have
.N()while other examples have.N.And while
.Nis one thing, there are a lot of use cases of subsetting with.SDand.I. Do we want to support both.SD()[var1 > 3, sum(var2)]in addition to.SD[var1 > 3, sum(var2)]?