fix(pandas): fast methods support DataFrame aggs like n()

[machow]

Supports the use of aggregation calls like n(_) inside filter, that go GroupBy[DataFrame] -> GroupByAgg. Most operations take a GroupBy[Series], so I wrongly reached for the Series.name attribute.

from siuba.data import mtcars
from siuba import _
from siuba.dply.vector import n
from siuba.experimental.pd_groups import fast_filter

# raised error about DataFrame not having a name attribute
mtcars >> fast_filter(n(_) > 1)

# worked okay
mtcars >> fast_filter(n(_.mpg) > 1)