Speed: why running in Python is MUCH slower as in GUI ???

[figoyouwei]

Dear Hines,

I am surprised, that no previous posts asked about this question ... but ...

I confirmed this quite a few times, trying to get rid of every possible loads on Python.

When running a small model, the time difference is tolerable, so I didn't notice that before ...
I just found that running a "not so big" model for 200 ms simulation time, takes about ~200 seconds in real time in GUI, with even the drawing in the graph.

But it takes about ~800 seconds in real time in Python.

The model is totally written in hoc, basically just loaded into Python. And I double-checked, e.g. no saving in Python, just run, to eliminate all sorts of speed factors, to make sure that the only difference in my case is the running environment ...

So have anybody else also encountered that ? and what is the explanation and possible solution ?

[ted]

To diagnose the cause of the problem, it will be necessary to reproduce it. Can you zip up "just enough code" to reproduce it (no binary files, no .o or .c or .dll files, no i686 directories) and email it to
ted dot carnevale at yale dot edu
?

And here are some very important questions:
Which version of NEURON are you using (what is the first line of the message that NEURON prints to its terminal when it starts?)?
What OS are you using?
What version and distribution of Python are you using?
And, finally,

The model is totally written in hoc, basically just loaded into Python.

How do you "load the model into Python"?

[figoyouwei]

It works normal again, when I put things in template ... which is even weirder ...
Let me diagnose it and confirm again myself before "wasting" your time on this.
thank you Ted for the reply.