cannot find -lpython2.3

[refriend]

I have been trying to execute several models from ModelDB but after each attempt of

Code: Select all

nrnivmodl

I get the error message

Code: Select all

 /usr/bin/ld: cannot find -lpython2.3
collect2: ld returned 1 exit status
make: *** [libnrnmech.la] Error 1

I am running nrn-6.1-rel1 on Fedora Release 7 (Moonshine). Python 2.5, numby and scipy are installed. Do i need to revert to an earlier version of Python?

I could not find the answer to this question via the forum or installation guides. I apologize if this has been previously answered.

Thanks!

[hines]

just today I uploaded
http://www.neuron.yale.edu/ftp/neuron/v ... 3.i686.rpm
which links against python2.5.
But let me know if you have problems because I installed python with --enable-shared so it looks for a python2.5.so...

From another point of view, it is not necessary for nrnivmodl to link against python since the nrnmech.so being constructed needs only neuron specific stuff from the neuron shared objects.
So it should be possible to edit
/usr/local/nrn/i686/bin/nrnmech_makefile
and comment out the PY_LIBS definition with a # and create an empty one that looks like

Code: Select all

PYLIBS =

[refriend]

Thanks!

I installed nrn-6.2.alpha-973.i686.rpm and successfully re-nrnivmodl a few models.

The following information may be more a result of the models or my nube NEURON skills rather than nrniv, but I thought I would pass on the info in case it was nrniv related. I've now tested a number of models and ALL of them report errors. I've include 2 examples but the others are similar.

1. Downloaded from: http://senselab.med.yale.edu/ModelDB/Sh ... odel=71312

Code: Select all

[refriend@localhost Ascol_ca1_2006]$ nrniv mosinit.hoc 
NEURON -- VERSION 6.2.973 (1919) 2007-12-16
Duke, Yale, and the BlueBrain Project -- Copyright 1984-2007
See http://www.neuron.yale.edu/credits.html

        1 
        0 
        0 
---------------------------Type I Model-----------------------------
        1 
        1 
        0 
total # of segments (100 hz): 912        max path distance:  726.352
        65 
nrniv: syntax error
 in ModelTypeI.hoc near line 130
                 insert nax gbar_nax=gna * AXONM        sh_nax=nash
                          ^
        xopen("ModelTypeI.hoc")
      xopen("fig4A.hoc")

2. Downloaded from: http://senselab.med.yale.edu/ModelDB/Sh ... odel=55035

Code: Select all

[refriend@localhost Migliore_ca1_2005]$ nrniv mosinit.hoc 
NEURON -- VERSION 6.2.973 (1919) 2007-12-16
Duke, Yale, and the BlueBrain Project -- Copyright 1984-2007
See http://www.neuron.yale.edu/credits.html

        1 
        1 
        1 
        1 
        0 
        1 
        1 
        1 
        1 
        1 
        1 
        0 
        1 
        1 
        0 
        0 
        0 
        0 
        1 
        0 
        1 
        1 
        1 
nrniv: syntax error
 in fig2A.hoc near line 88
                 insert nax gbar_nax=gna * AXONM
                          ^
        xopen("fig2A.hoc")
      execute1("{xopen("fig2A.hoc")}")
    load_file("fig2A.hoc")
        0 
nrniv: syntax error
 in fig2A.hoc near line 89
                 insert kdr gkdrbar_kdr=gkdr
                          ^

[hines]

$ nrniv mosinit.hoc
...
nrniv: syntax error
in ModelTypeI.hoc near line 130
insert nax gbar_nax=gna * AXONM sh_nax=nash

Under linux, the compiled mod files are not loaded when nrniv is directly launched
(the exception is when launching with the
-mpi option when running under mpiexec.)
Instead, one typically launches using
nrngui mosinit.hoc
which is a script that launches the
i686/special script which launches
nrniv with the proper -dll argument that dynamically loads i686/.libs/nrnmech.so

You can tell you have loaded the mod files because they are listed just after the banner. As in

Code: Select all

[hines@localhost bulb1]$ nrngui
NEURON -- VERSION 6.2.972 (1918) 2007-12-15 (1918M)
Duke, Yale, and the BlueBrain Project -- Copyright 1984-2007
See http://www.neuron.yale.edu/credits.html

loading membrane mechanisms from /home/hines/bulb1/x86_64/.libs/libnrnmech.so
Additional mechanisms from files
 ampanmda.mod fi.mod kamt.mod kdrmt.mod naxn.mod ThreshDetect.mod
oc>

[refriend]

You are absolutely correct. So it turns out to have been a nube mistake (I kind of saw that coming).

Thanks for the quick responses.