I have the following problem:
I applying to a single-cell model the Multisplit parallelization.
I know that one of the key steps is to put in the code function "section_exists()" to manage the access to the sections assigned to different hosts.
My approach is to run the code and correct it with section_exists() when at the points where occur the error like "section was deleted".
Now, I have the following error:
Code: Select all
NMDA[79] point process not located in a section
in init_Ca1-newPoirazi_181110_syn.hoc near line 57
1 dorun()
^
1 NetCon(NetStim[79], NMDA[79])
1 stimul_syn(238.8, 199)
1 run_program(199, 238.8, 0, 2, "Control_")
1 init()
I tried to print the name of the sections involved with "print secname()" and I have seen that these are "soma(0)" and "apical_dendrite(64)".
Below, I show the code of the procedure that causes the error:
Code: Select all
proc stimul_syn() { local r,s localobj nc,syn,src
nclist1=new List()
nclist2=new List()
for r=0, $1-1{ //$1 n_NMDA
stim[r].start=rsstim.repick()
src = stim[r]
print r, secname()
syn = syn1list.object(r)
nc = new NetCon(src,syn)
nc.delay=0.1
nc.weight=1
nclist1.append(nc)
}
for s=0, $2-1{
stim[s+$1].start=rsstim.repick()
src = stim[s+$1] // $2 nsyn
print s, secname()
syn = syn3list.object(s)
nc = new NetCon(src,syn)
nc.delay=0.1
nc.weight=1
nclist2.append(nc)
}
}
Also, when I run with a single host (in the previous case hosts were 2), if I print the secname, section is always the same: soma(0).
I apologize if I was not very clear in exposing the problem.
Thanks, Pasquale.