Comparison of identical Ding0 data does not always confirm equality

[gplssm]

The function for comparison of Ding0 networks sometimes state branch IDs aren't the same.
This can be reproduced by comparing 5-20 datasets generated by exactly the same Ding0 code and input data. The result is like this. The grid is most probably identical, but for some reason the IDs do not match.

Hints

• This issue is maybe simply related to different sorting while ID setting in Ding0 and sorting in the comparison function.
• Both, the branch setting method and to_dataframe use graph_edges that return a sorted list of edges (sorted by nodes).

Tasks

• Make sure Ding0 is deterministic regarding IDs of branches
• Revise comparison function to detect equality properly
• Maybe comparison function such that it always returns more information about inequality

[nesnoj]

According to #97, this was not completely solved. Refer to #40 (comment) for further details and proposed investigations.

[mltja]

Already fixed