Discussion on best way to visualise data in an experiment

At the hackathon, there was a lot of discussion on the importance of data visualisation. There are already a lot of nice tools for visualising data in spikeinterface (plot_traces(), and spikeinterface-gui). These are very useful for having a good look, close up at a single dataset.

However, when running a lot of acquisition sessions, it can be useful to print summary of a recording to quickly assess data quality across an entire experiment. I was wondering how people are handling this in their research? I sometimes use code blocks like the below (psedocode) to dump images to a PDF for review later. An intrinsic problem at what scale to plot the data - at large time bins you loose all spike resolution. At small time bins, you need to take small snapshots through a long recording. In this case it might be easy to miss large blocks of the recording where data quality could feasibly be comprimised. Also, if saving to .svg / .png you cant really zoom in like a usual matplotlib recording.

from matplotlib.backends.backend_pdf import PdfPages
import spikeinterface.widgets as si_widgets

pdf_pages = PdfPages(output_filepath)
fig, ax = plt.subplots(*(1, 3))

# some list of times to plot, with a period (i.e. window) to plot

for i, (start_time, stop_time) in enumerate(
        zip(times,
            times + period_s)
 ):

        si_widgets.plot_traces(
            recording,
            order_channel_by_depth=True,
            time_range=(start_time, stop_time),
            mode="line",
            show_channel_ids=True,
            return_scaled=True,
            ax=ax[i],
        )
        ax[i].get_legend().remove()

pdf_pages.savefig(fig)
plt.close(fig)
pdf_pages.close()

Start by reviewing the existing plot_traces() workflow and the spikeinterface-gui tool, then compare them with the provided matplotlib/PDF example. The issue does not define a specific implementation or acceptance criteria, so the visualization requirements and what constitutes a completed summary need to be clarified first.