Shanks, probes and configurations

The GUI can show several recordings at once. This page explains the three ways it groups them, how they relate to each other, and why the choice matters for where the results are written.

Shanks

A shank is the unit the GUI aligns. Every shank has its own traced trajectory, its own reference lines and its own fit, even when several shanks belong to one probe. A multi-shank probe is therefore aligned once per shank.

Showing the shanks of a probe side by side still helps: the shanks are fixed relative to each other, so the alignment of one is a check on its neighbours. A boundary that one shank crosses at a given depth should appear at a similar depth on the shank next to it, if the anatomy is similar there.

Probes in the session YAML

A probe entry in the session YAML is a set of dataset folders: spike sorting, raw data, picks and output. How a multi-shank probe is entered depends on how it was spike sorted.

Sorted as a whole

One probe entry. The spike sorting holds every shank, and the GUI splits the channels into shanks using the probe geometry (see Probe geometry). It expects one traced trajectory per shank, xyz_picks_shank1.json, xyz_picks_shank2.json and so on, and writes the results with the same shank suffix. All shanks can share one output folder.

Sorted one shank at a time

One probe entry per shank. Each entry is a single-shank probe as far as the GUI is concerned: it expects xyz_picks.json and writes channel_locations.json with no suffix. Two entries writing to the same output folder would overwrite each other’s results and read each other’s previous alignments, so each needs its own output folder.

Either way the shanks look the same in the GUI: one tab, or one column in the grid layout, per shank.

Configurations

A configuration is a complete set of probes. With two configurations, the GUI shows the same probes from two different sources side by side, for example:

• the same recording spike sorted with two sorters or with different parameters

• the same probe recorded with two channel maps, such as dense and sparse

• the same data before and after a change to the preprocessing

The two configurations are aligned together: a reference line placed on one applies to the same shank in the other, since they are the same piece of tissue. This is why both configurations must contain the same probe names; the names are how the GUI matches them up. At most two configurations can be loaded.

Choosing a layout

Your data	Session YAML
One single-shank probe	One probe entry
One multi-shank probe, sorted together	One probe entry, one picks file per shank
One multi-shank probe, sorted per shank	One probe entry per shank, each with its own output
Two versions of the same recording	Two configurations with the same probe names

The keys are listed in Session YAML.