/*
* test_iaf_psp.sli
*
* This file is part of NEST.
*
* Copyright (C) 2004 The NEST Initiative
*
* NEST is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 2 of the License, or
* (at your option) any later version.
*
* NEST is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with NEST. If not, see <http://www.gnu.org/licenses/>.
*
*/
/* BeginDocumentation
Name: testsuite::test_iaf_psp - sli script for test of iaf_neuron spike input
Synopsis: (test_iaf_psp) run -> compare response with reference data
Description:
test_iaf_psp.sli checks the voltage response of the iaf_neuron
model neuron to a single incoming spike. The voltage excursion is
called post-synaptic potential (PSP). In the iaf_neuron model neuron
the post-synaptic current is described by an alpha-function
(see [1] and references therein). The resulting PSP has a finite
rise-time, with voltage and current beeing zero in the initial
condition (see [1]).
The dynamics is tested by connecting a device that emits spikes
at individually configurable times (see test_spike_generator) to
a model neuron.
The weight of the connection specifies the peak value (amplitude)
of the post-synaptic current (PSC) in pA.
The subthreshold dynamics of the iaf_neuron is integrated exactly.
Therefore, it is suitable to check whether the simulation kernel
produces results independent of the computation step size
(resolution).
In order to obtain identical results for different computation
step sizes h, the SLI script needs to be independent of h.
This is achieved by specifying all time parameters in milliseconds
(ms). In particular the time of spike emission and the synaptic
delay need to be integer multiples of the computation step sizes
to be tested. test_iaf_dc_aligned_delay demonstrates the strategy
for the case of DC current input.
The expected output is documented and briefly commented at the end of
the script.
References:
[1] Rotter S & Diesmann M (1999) Exact simulation of time-invariant linear
systems with applications to neuronal modeling. Biologial Cybernetics
81:381-402.
Author: July 2004, Diesmann
SeeAlso: testsuite::test_iaf, testsuite::test_iaf_dc_aligned_delay, testsuite::test_spike_generator
*/
/unittest (6688) require
/unittest using
1.0 /delay Set % in ms
/AlignedImpact
{
<< >> begin
/d Set /h Set
ResetKernel
0 <<
/local_num_threads 1
/resolution h
>> SetStatus
/spike_generator Create /sg Set
sg <<
/precise_times false
/origin 0.0 % in ms
/spike_times [ 2.0 ] % in ms
/start 1.0 % in ms
/stop 3.0 % in ms
>> SetStatus
/iaf_neuron Create /neuron Set
/voltmeter Create /vm Set
vm << /withtime true /to_memory true /time_in_steps true /interval h >> SetStatus
/spike_detector Create /sd Set
sd << /withtime true /to_memory true /time_in_steps true >> SetStatus
sg neuron 100.0 delay Connect
vm neuron Connect
neuron sd Connect
7.0 Simulate
d Transpose First /test_times Set % times of reference
vm [/events [/times /V_m]] get cva % array of recorded voltages
6 ToUnitTestPrecision % to precision of reference
Transpose % all recorded tuples
{First test_times exch MemberQ } Select % those with reference
d eq % compare
end
} def
{
% h d, vector of resolutions and compressed reference data
InflateUnitTestData
Transpose {AlignedImpact} MapThread
true exch {and} Fold % remove this line to see individual results
}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%
% Expected output of this program:
%
% The output send to std::cout is a superposition of the output of
% the voltmeter and the spike detector. Both, voltmeter and spike
% detector are connected to the same neuron.
%
%
% h= (in ms)
[ 0.1 0.2 0.5 1.0 ]
%
% time voltage
%
[
% 0 0 0 0 -70
[ 1 -70]
[ 2 1 -70]
[ 3 -70]
[ 4 2 -70]
[ 5 1 -70]
[ 6 3 -70]
[ 7 -70]
[ 8 4 -70]
[ 9 -70]
[ 10 5 2 1 -70]
[ 11 -70]
[ 12 6 -70]
[ 13 -70]
[ 14 7 -70]
[ 15 3 -70]
[ 16 8 -70]
[ 17 -70]
[ 18 9 -70]
[ 19 -70]
[ 20 10 4 2 -70]% <-- Spike generator emits spike
[ 21 -70]% with time stamp t=2.0 ms.
[ 22 11 -70]
[ 23 -70]
[ 24 12 -70]% <-- Synaptic delay of 1.0 ms.
[ 25 5 -70]
[ 26 13 -70]
[ 27 -70]
[ 28 14 -70]
[ 29 -70]
[ 30 15 6 3 -70]% <--------- Spike arrives at neuron and
[ 31 -69.9974]% <- changes the state vector of the
[ 32 16 -69.9899]% | neuron, not visible in voltage
[ 33 -69.9781]% | because voltage of PSP initial
[ 34 17 -69.9624]% | condition is 0.
[ 35 7 -69.9434]% |
[ 36 18 -69.9213]% -- Arbitrarily close to the time
[ 37 -69.8967]% of impact (t=3.0 ms) the effect
[ 38 19 -69.8699]% of the spike (PSP) is visible in
[ 39 -69.8411]% the voltage trace.
[ 40 20 8 4 -69.8108]%
[ 41 -69.779 ]
[ 42 21 -69.7463]
[ 43 -69.7126]
[ 44 22 -69.6783]
[ 45 9 -69.6435]
[ 46 23 -69.6084]
[ 47 -69.5732]
[ 48 24 -69.538 ]
[ 49 -69.5029]
[ 50 25 10 5 -69.4681]% <--- The voltage trace is independent
[ 51 -69.4336]% of the computation step size h.
[ 52 26 -69.3995]% Larger step sizes only have fewer
[ 53 -69.366 ]% sample points.
[ 54 27 -69.333 ]
[ 55 11 -69.3008]
[ 56 28 -69.2692]
[ 57 -69.2383]
% ...
]
%
rolld exec assert_or_die