www.neuron.yale.edu

Hello everyone!

I'm trying to launch a Sonata network in NEURON. According to their documentation they store synapses locations as some section id. I suspect that it is a Section's index in h.allsec() collection. I've looked over Section's documentation and have some questions.
1. Am I correct that there is no such thing as Section's id? At least publicly available in NEURON API.
2. Is there a shortcut method to get a section by its index? Currently I transform h.allsec() into a list, and access the necessary section by the list's index.
3. I've tried to create a synapse by h.Exp2Syn(sec=section_id) but it didn't work. However in this documentation a method h.this_section is mentioned that allows to get something that looks like Section's id.

Internally, the Sections are stored in a linked list. There is thus no direct, consistent-across-runs section indexing scheme... but as you've observed, there's no advantage for NEURON storing such a list since you can create it quickly in the non-parallel case:
Once you have that, you are free to use that to position synapses, e.g. the following code places a synapse at the midpoint of the ith entry in seclist:
A few comments:

• Don't use this_section. It's not guaranteed to work on 64-bit machines, which nowadays is all machines.
• Remember, Python semantics don't copy the sections when creating a list, they only store a reference to the section, so constructing seclist above is fast.
• Synapses are positioned at normalized locations within a Section (remember: there can be many segments associated with a given section ).
• Sections have name and cell attributes that can be used to provide unique identifiers.
• A seclist created in the above way would not be suitable for use in parallel simulation since different machines would have different section 0s. You'd have to work with the name/cell (or ideally through a cell class) to uniquely identify sections in a parallel environment.

Thank you for such a thorough answer! By any chance, do you have an example of identifying sections in parallel environment?