nichesphere.niche_stats.community_edge_weight_distributions

nichesphere.niche_stats.community_edge_weight_distributions(g: Graph, partition: dict) → dict

Raw edge weight lists per community, split by location and sign.

Uses the same unique-pair iteration as community_edge_stats() so weights correspond one-to-one with the counts computed there. External edges are attributed to both communities they connect.

Parameters:

• g (nx.Graph) – Signed co-localization network.

• partition (dict) – Mapping of node → community name.

Returns:

dist – Keyed by community name; each value is a dict with four lists:

• 'internal_pos' – positive weights of internal edges

• 'internal_neg' – absolute values of negative internal edges

• 'external_pos' – positive weights of external edges

• 'external_neg' – absolute values of negative external edges

Return type:

dict

Examples

>>> dist = nichesphere.niche.community_edge_weight_distributions(gCol, vc_map_named)
>>> plt.hist(dist['2_FibCore']['internal_pos'])