www.neuron.yale.edu

Hi,

I built a complicated morphology. I would like to be able to duplicate many clones of that morphology and then wire them together using the usual "connect" statement. I figured the way to do it is to embed the hoc code in a template and then generate as many instances of the morphology as I need. My question is: what is the correct syntax or method for then directly connecting them in this way? I naively (not having used templates before) tried something like "connect Cell[1].dend(0), Cell[0].dend(1)" with dend made public and all, but that doesn't do it.

Thanks

Can't tell where your code went wrong without seeing it, but I'll email you a little program I wrote a while back that may help you find your error and maybe give you some new ideas.

Better than email: pick up this file
http://www.neuron.yale.edu/ftp/ted/neuron/btree6s.hoc
and use NEURON to execute it. It could probably be more elegant, but it should at least get you going.

Ted, it works when I follow the example in your code and connect the "template tree" to a section created outside of the template but not when connecting two template trees.

In other words, from your code: is fine, but: just gives an error.

Is it necessary to connect two template trees through an intermediate section?

Tell me if the error message you saw is something like this: This is somewhat puzzling because the error is localized to the second occurrence of [...](...) on the line, and I would have expected hoc to gag on the first occurrence. But that's the way it works.

So the error message implies that the parent section must be specified in a different way. The documentation of the connect statement presents these two forms:
connect childsection(0or1), x
connect childsection(0or1), parent(x)

Try the first form, and it works--e.g. treats stree[0].dend[0] as the parent section, stree[1].dend[0] as the child, and attaches the 0 end of the child to the 1 end of the parent.

I got the same error message as you, but using the first form of the connect statement, as you suggest, does the trick.

Many thanks for the help! I suspect I would have beaten my head against the keyboard for a long time before thinking the connect statement would work in one form but not the other.