Example given `d = mdf.MicroDataFrame(dict(g=["a", "a", "b"], y=[1, 2, 3]), weights=[4, 5, 6])`: `d.groupby("g").sum()` works: ``` y -- 14.0 18.0 ``` `d.groupby("g").y.sum()` also works: ``` g a 14.0 b 18.0 dtype: float64 ``` `d.groupby("g").sum()["y"]` works too: ``` g a 14.0 b 18.0 Name: y, dtype: float64 ``` `d.groupby("g")[["y"]].sum()` does not: ``` | y -- | -- 3 3 ``` `d.groupby("g")["y"].sum()` also doesn't work: ``` g a 3 b 3 Name: y, dtype: int64 ``` `d.groupby("g").sum()[["y"]]` produces a `KeyError` (see #192).