Bug Fix: Invalid group detected for dtype = "boolean" when plotting

[tomcarter23]

When using the attr_distr_plot function the following error was encountered:

File ~/venv/lib/python3.8/site-packages/fairlens/plot/distr.py:77, in distr_plot(df, target_attr, groups, distr_type, show_hist, show_curve, shade, normalize, cmap, ax)
     74 if ax is None:
     75     ax = plt.gca()
---> 77 preds = utils.get_predicates_mult(df, groups)
     79 cmap = cmap or sns.color_palette("deep")
     80 palette = itertools.cycle(cmap)

File ~/venv/lib/python3.8/site-packages/fairlens/utils.py:318, in get_predicates_mult(df, groups)
    316     else:
    317         if not isinstance(group, pd.Series) or group.dtype != "bool":
--> 318             raise ValueError(
    319                 "Invalid group detected. Groups must be either dictionaries or pandas series' of bools."
    320             )
    322         predicates[i] = group
    324 # Check all attributes are valid

ValueError: Invalid group detected. Groups must be either dictionaries or pandas series' of bools.

This was because the groups were created with type "boolean" which incorrectly fails the comparison group.dtype != "bool". Using the pandas is_bool_dtype avoids this