Newly proposed upstream rule: constant column check with nunique ```python series.nunique() == 1 ``` Rather use (in the absence of nans): ```python v = series.values (v[0] == v).all() ``` See https://github.com/deppen8/pandas-vet/issues/119