Problem running benchmarks from Tikidji-Hamburyan et al 2017

Support
1

I tried to run following a model file at the following link ModelDB: Brain networks simulators - a comparative study (Tikidji-Hamburyan et al 2017), but the result was an error (OSError: …/ input.ssv not found). Can you explain it so that my running results with the one in the file can produce the same result or not an error?

import time
import numpy as np
import matplotlib
matplotlib.rcParams[“savefig.directory”] = “”
from brian2 import *

startbuild=time.time()

Ne,Ni,Ns=4000,1000,500

excw,excd = 0.009, 0.8
inhw,inhd = -0.050, 2.1
stmw,stmd = 0.025, 0.5
inspk = np.genfromtxt("…/input.ssv")

tau = 10*ms
eqs = ‘’’
dv/dt = -v/tau : 1 (unless refractory)
‘’’

defaultclock.dt = 0.1*ms

E = NeuronGroup(Ne, eqs, threshold=‘v>1.’, reset=‘v = 0’, method=‘linear’, refractory=5.01ms)
E.v = 0
I = NeuronGroup(Ni, eqs, threshold=‘v>1.’, reset=‘v = 0’, method=‘linear’, refractory=5.01ms)
I.v = 0
S = SpikeGeneratorGroup(Ns, inspk[:,1].astype(int), inspk[:,0]*ms)

Cee = Synapses(E, E, on_pre='v_post += excw ')
prepost = np.genfromtxt("…/ee.ssv").astype(int)
Cee.connect(i=prepost[:,0], j=prepost[:,1])
Cee.delay = excd*ms

Cei = Synapses(E, I, on_pre='v_post += excw ')
prepost = np.genfromtxt("…/ei.ssv").astype(int)
Cei.connect(i=prepost[:,0], j=prepost[:,1])
Cei.delay = excd*ms

Cie = Synapses(I, E, on_pre='v_post += inhw ')
prepost = np.genfromtxt("…/ie.ssv").astype(int)
Cie.connect(i=prepost[:,0], j=prepost[:,1])
Cie.delay = inhd*ms

Cii = Synapses(I, I, on_pre='v_post += inhw ')
prepost = np.genfromtxt("…/ii.ssv").astype(int)
Cii.connect(i=prepost[:,0], j=prepost[:,1])
Cii.delay = inhd*ms

Cse = Synapses(S, E, on_pre='v_post += stmw ')
prepost = np.genfromtxt("…/se.ssv").astype(int)
Cse.connect(i=prepost[:,0], j=prepost[:,1])
Cse.delay = stmd*ms

e_mon = SpikeMonitor(E)
i_mon = SpikeMonitor(I)
s_mon = SpikeMonitor(S)

endbuild=time.time()

run(1 * second)

endsimulate= time.time()

print(“Building time : %.2f s”%(endbuild-startbuild ))
print(“Simulation time : %.2f s”%(endsimulate-endbuild))
print(“Time step : %.2f ms”%(defaultclock.dt*1000.))

e_mon = dstack( (np.array(e_mon.t/ms),np.array(e_mon.i)) )[0]
i_mon = dstack( (np.array(i_mon.t/ms),np.array(i_mon.i)) )[0]
s_mon = dstack( (np.array(s_mon.t/ms),np.array(s_mon.i)) )[0]

e_mon = e_mon[np.where(e_mon[:,0]<200)]
i_mon = i_mon[np.where(i_mon[:,0]<200)]
s_mon = s_mon[np.where(s_mon[:,0]<200)]

plot(e_mon[:,0], e_mon[:,1] , ‘.k’)
plot(i_mon[:,0], i_mon[:,1]+Ne+100 , ‘.r’)
plot(s_mon[:,0], s_mon[:,1]+Ne+Ni+200, ‘.b’)

show()
OSError Traceback (most recent call last)
in
6 inhw,inhd = -0.050, 2.1
7 stmw,stmd = 0.025, 0.5
----> 8 inspk = np.genfromtxt("…/input.ssv")
9
10 tau = 10*ms

C:\ProgramData\Anaconda3\lib\site-packages\numpy\lib\npyio.py in genfromtxt(fname, dtype, comments, delimiter, skip_header, skip_footer, converters, missing_values, filling_values, usecols, names, excludelist, deletechars, replace_space, autostrip, case_sensitive, defaultfmt, unpack, usemask, loose, invalid_raise, max_rows, encoding)
1747 fname = os_fspath(fname)
1748 if isinstance(fname, str):
→ 1749 fid = np.lib._datasource.open(fname, ‘rt’, encoding=encoding)
1750 fid_ctx = contextlib.closing(fid)
1751 else:

C:\ProgramData\Anaconda3\lib\site-packages\numpy\lib_datasource.py in open(path, mode, destpath, encoding, newline)
193
194 ds = DataSource(destpath)
→ 195 return ds.open(path, mode, encoding=encoding, newline=newline)
196
197

C:\ProgramData\Anaconda3\lib\site-packages\numpy\lib_datasource.py in open(self, path, mode, encoding, newline)
533 encoding=encoding, newline=newline)
534 else:
→ 535 raise IOError("%s not found." % path)
536
537

OSError: …/input.ssv not found.

2

Hi and welcome to the Brian forum. In general, please do not ask questions by replying to unrelated topics, instead open a new topic (I moved your post to a new topic). Please also note that this forum is mainly for questions related to Brian itself, not about code from others that use Brian. In such cases, it is usually best to contact the others of the code in question directly.
Having said all that, I think the problem in your case is simply that you try to run the example.py file from ModelDB in isolation, but it needs to load other files to work correctly. You will either have to download the full project as a zip file and unpack it, or clone the full project from github. You can then run the file from its directory, and it should be able to find the additional files it needs in the parent directory.

1 Like
3

Hi @nadi,
@mstimberg is completely right. These tests were designed to have the same input files for all simulators. Therefore, they are located in the test directory, while all simulator-specific scripts are in local directories. The directory tree is that:

test directory
├── BRIAN
│   └── example.py
├── NEST
│   └── example.py
├── NEURON
│   ├── example.py
│   └── x86_64
│       └── (module binary)
├── ee.ssv
├── ei.ssv
├── ie.ssv
├── ii.ssv
├── input.ssv
└── se.ssv

You need to reconstruct this directory structure exactly to make scripts work. The simplest and the best way to achieve that is to download the whole ModelDB zip archive and unpack it (as @mstimberg suggested).

1 Like
4

Ok thank you @mstimberg
I Will to try again

Pada tanggal Sel, 27 Apr 2021 00.32, Ruben Tikidij-Hamburyan via Brian simulator <brian@discoursemail.com> menulis: