Command: testsuite::test_iaf_psp_normalized

Synopsis

(test_iaf_psp_normalized) run -> compare response with desired outcome

[1] Rotter S & Diesmann M (1999) Exact simulation of time-invariant linear

systems with applications to neuronal modeling. Biologial Cybernetics

81:381-402.

[2] Galassi M. Davies J. Theiler J. Gough B. Jungman G. Booth M.

& Rossi F. (2006). GNU Scientific Library Reference Manual (2nd Ed.).

Network Theory Limited.

The script computes the peak location of the PSP analytically for a

neuron model with an alpha-shaped post-synaptic current (PSC) [1]. In case

the GNU Scientific Library (GSL) is not present the peak location is

found by searching for the root of the derivative of the PSP. We then

compute the peak value for a PSC with unit amplitude and show how the

synaptic weight can be adjusted to cause a PSP of a specific

amplitiude. Finally we check whether the simulation indeed generates

a PSP of the desired amplitude.

In application code the test for the availability of the GSL is not

necessary because NEST has a built in version of the LambertWm1 which

automatically replaces the GSL function if required. This removes the

need to specify the derivative of the function of interest here the

PSP in application code. A further alternative is used in

test_lambertw where knowledge of the range of values of the

non-principal branch of the Lambert-W function [-1 -\infty) is

exploited to find the inverse of x*exp(x) by bisectioning.

References

[1] Rotter S & Diesmann M (1999) Exact simulation of time-invariant linear

systems with applications to neuronal modeling. Biologial Cybernetics

81:381-402.

[2] Galassi M. Davies J. Theiler J. Gough B. Jungman G. Booth M.

& Rossi F. (2006). GNU Scientific Library Reference Manual (2nd Ed.).

Network Theory Limited.

Description

The script computes the peak location of the PSP analytically for a

neuron model with an alpha-shaped post-synaptic current (PSC) [1]. In case

the GNU Scientific Library (GSL) is not present the peak location is

found by searching for the root of the derivative of the PSP. We then

compute the peak value for a PSC with unit amplitude and show how the

synaptic weight can be adjusted to cause a PSP of a specific

amplitiude. Finally we check whether the simulation indeed generates

a PSP of the desired amplitude.

In application code the test for the availability of the GSL is not

necessary because NEST has a built in version of the LambertWm1 which

automatically replaces the GSL function if required. This removes the

need to specify the derivative of the function of interest here the

PSP in application code. A further alternative is used in

test_lambertw where knowledge of the range of values of the

non-principal branch of the Lambert-W function [-1 -\infty) is

exploited to find the inverse of x*exp(x) by bisectioning.

File

testsuite/unittests/test_iaf_psp_normalized.sli

Author

July 2009
Diesmann