Jonas I. Liechti

It might make sense to split up src/flowstab/flow_stability.py into a clustering and into a analysis management part. The sub-module is already quite large.

We should provide a single method to preforms all the steps, up to find_luvain_clustering form FlowIntegralClustering. A compact way of running an analysis would then look like this: - create...

@alexbovet The interface class is functional and the documentation is adapted: [README.md](https://github.com/alexbovet/flow_stability/tree/53-unified-interfacing-through-flowstability-class?tab=readme-ov-file#usage) Then the docs: [quickstart](https://flow-stability--59.org.readthedocs.build/en/59/quickstart.html) [flowstab module](https://flow-stability--59.org.readthedocs.build/en/59/autoapi/flowstab/index.html) [flow_stability sub-module](https://flow-stability--59.org.readthedocs.build/en/59/autoapi/flowstab/flow_stability/index.html) Should we merge this?

First we need an example dataset, some minimal test data to construct a temporal network.

@alexbovet: What is the rationale behind allowing the setting of [`node_to_label_dict`](https://github.com/alexbovet/flow_stability/blob/85ae8c597b267da65d193a4cfe13ec0412c04ea8/src/flowstab/temporal_network.py#L99) when initializing a `ContTempNetwork`? I understand that this attribute is beneficial as it facilitates the mapping of node IDs...

> Yes, this was probably because I needed it at some point. You are proposing to relabel nodes by default and dropping the custom node_to_label_dict? I'm fine with it. Great!...

@alexbovet: is [`fix_tau_k`](https://github.com/alexbovet/flow_stability/blob/85ae8c597b267da65d193a4cfe13ec0412c04ea8/src/flowstab/temporal_network.py#L1322) used for anything other than handling instantaneous events? From what I see it is only used in case we have instantaneous events. If it is only for...

@alexbovet for instantaneous events, is there a reason why the event duration is given by $`t_{i+1} - t_{i}`$ of unique starting times: https://github.com/alexbovet/flow_stability/blob/85ae8c597b267da65d193a4cfe13ec0412c04ea8/src/flowstab/temporal_network.py#L1925-L1926 and not just $`1`$, as is the...

@alexbovet for the `ContTempInstNetwork` class, the laplaican computation is done differently. However, temporal networks with instantaneous events are treated like regular temporal networks with events durations = 1, right?! Why...

@alexbovet in the instantaneous events case we set the event durations to `1`. However, when we compute the inter event transitions matrices, $`\tau_k`$ is given by the inter-event time, which...