'LinearSegmentedColormap' object has no attribute 'colors' with divergent network edges

[eilidhmacnicol]

• nxviz version: 0.7.3
• Python version: 3.9.7
• Operating System: MacOS 12.3.1

Description

Adding colour map to a divergent network crashes, but not for the same network if only positive or negative connections are used.

What I Did

Command: edge_color = edges.edge_colors(et, nt=None, color_by='weight', node_color_by=None)

Traceback:

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
/var/folders/4v/g8pg_7r950bf7vtkg09mpc240000gn/T/ipykernel_9283/651244063.py in <module>
----> 1 edge_color = edges.edge_colors(et, nt=None, color_by='weight', node_color_by=None)

~/miniconda3/envs/analysis/lib/python3.9/site-packages/nxviz/edges.py in edge_colors(et, nt, color_by, node_color_by)
     54         )
     55     elif color_by:
---> 56         return encodings.data_color(et[color_by], et[color_by])
     57     return pd.Series(["black"] * len(et), name="color_by")
     58 

~/miniconda3/envs/analysis/lib/python3.9/site-packages/nxviz/encodings.py in data_color(data, ref_data)
    121     """
    122     cfunc = color_func(ref_data)
--> 123     return data.apply(cfunc)
    124 
    125 

~/miniconda3/envs/analysis/lib/python3.9/site-packages/pandas/core/series.py in apply(self, func, convert_dtype, args, **kwargs)
   4354         dtype: float64
   4355         """
-> 4356         return SeriesApply(self, func, convert_dtype, args, kwargs).apply()
   4357 
   4358     def _reduce(

~/miniconda3/envs/analysis/lib/python3.9/site-packages/pandas/core/apply.py in apply(self)
   1034             return self.apply_str()
   1035 
-> 1036         return self.apply_standard()
   1037 
   1038     def agg(self):

~/miniconda3/envs/analysis/lib/python3.9/site-packages/pandas/core/apply.py in apply_standard(self)
   1090                 # List[Union[Callable[..., Any], str]]]]]"; expected
   1091                 # "Callable[[Any], Any]"
-> 1092                 mapped = lib.map_infer(
   1093                     values,
   1094                     f,  # type: ignore[arg-type]

~/miniconda3/envs/analysis/lib/python3.9/site-packages/pandas/_libs/lib.pyx in pandas._libs.lib.map_infer()

~/miniconda3/envs/analysis/lib/python3.9/site-packages/nxviz/encodings.py in discrete_color_func(val, cmap, data)
     75     """
     76     colors = sorted(data.unique())
---> 77     return cmap.colors[colors.index(val)]
     78 
     79 

AttributeError: 'LinearSegmentedColormap' object has no attribute 'colors'

Continuous connection weights (positive and negative alone) runs as expected.

[eilidhmacnicol]

I think the problem is that divergent_color_func is defined but never called, so I made a small change that seems to solve the problem here: 4ec28c3

Let me know if this is likely the case and I can make a pull request