testsuite::test_iaf_dc_aligned_automatic

Command: testsuite::test_iaf_dc_aligned_automatic


Synopsis
(test_iaf_dc_aligned_automatic) run -> compare response with reference data


Description


test_iaf_dc_aligned_automatic.sli is a modified version of
test_iaf_dc_aligned_delay.sli .

The code of this program is now identical to
test_iaf_dc_aligned_delay.sli. Previous versions used a different
definition of start which was not independent of the resolution.
Therefore an additional device property /on was required. See
test_iaf_dc_aligned_automatic.sli for details.

The documentation below refers to the old definition of start

The output generated is exactly identical to the output of the
previous file. The dependence of the simulation script on the
computation step size is completely removed by using the capability
of the DC generator to adjust the required starting time to the
computation time step. This functionality is implemented in the Device
base class and thus available to all devices.

Instead of specifying the start time of the generator as in
test_iaf_dc_aligned_delay.sli the emission time of the first current
event is specified. A device can only take action in a step t-h -> t.
When t-h represents the start time of the device t is the earliest time
stamp of an emitted event. In our case the time stamp is the onset of the
current. Thus both times can not simultaneously be independent of the
computation time step h. It is sufficient to specify one of the two measures
the other one is computed automatically. Internally only the start time
is stored. A warning is issued when both measures are specified because
a difference other than h would be meaningless and a difference of h may
not be representable by the two double values. In case both properties are
specified on overrides start. The timer of the device does not rely on
computing differences between doubles but uses the exact representation of
time provided by the simulation kernel.

An inspection of the property dictionary of the DC generator at different
computation step sizes shows the relationship between DC generator
on and start.

h=0.1ms h=0.5ms
<< <<
/start 1.9 /start 1.5
/on 2.0 /on 2.0
>> >>

No output of the spike detector is visible here because of the short simulation
time of only 6 ms. Set the simulation time to larger values to verify the
consistency of voltage meter and spike detector.

The expected output is documented at the end of the script.

July 2004
updated Diesmann

File
testsuite/unittests/test_iaf_dc_aligned_automatic.sli