In probeinterface, they have a nice plotting library and allow you to plot the probe.

It would be nice to:

1. Expose a function that displays a plot of the probe (uses probeinterface plot_probe at the run level
2. Automatically save this to the run folder when save_preprocessing is run

It is hard to decide whether to save this at the session level or the run level. The probes should be the same across all runs within a session. However, to add flexibility and avoid edge-cases, it makes sense to save this as a per-run output, even if it will duplicate a bit. However, a session.plot_probe() function should also be exposed to quickly plot the probe associated with the session.

The implementation will be something like:

1. on _preprocess_run, add a function called 'save_probe_plot'. I am not actually sure how plot_probe will look when the recording is split per-shank, will need to investigate. The spikeinterface recording objects have a function called get_probe().
2. on session, a function like plot_probe(). This will have to check all probes attached to all runs and ensure they are the same, then it can plot one of them.

The code organization will roughly match plot_preprocessed approach.