Cycle enumeration strategy in `get_cycle_closure_dataframe`

[hannahbaumann]

In get_cycle_closure_dataframe() we currently enumerate cycles using:

cycles = [
    c for c in nx.simple_cycles(network.to_undirected())
    if len(c) <= max_cycle_length
]

During review we discussed whether nx.cycle_basis() would be more appropriate than nx.simple_cycles().

The main question is whether cycle closure analysis should use:

• all simple cycles (simple_cycles, including “super cycles”), or
• only a minimal independent basis (cycle_basis).

For now, we decided to keep simple_cycles + max_cycle_length.

Reasoning:

• Larger/super cycles may help avoid missing problematic edges due to cancellation in smaller cycles.
• Downstream edge-level diagnostics (get_cycle_closure_edge_statistics_dataframe) benefit from evaluating edges across multiple cycles.
• This aligns with observations that closing basis cycles does not necessarily imply larger cycles will also close (e.g. DOI: 10.1021/acs.jcim.5c00554).

We should revisit this if super cycles appear noisy or uninformative.