Critical behaviour in NetCon.delay

[URig]

While fitting parameters of a synapse (AMPA_S from the Neuron Book)
and its NetCon-object critical sensitivity on the value of delay is
reproducible observed.

Three parameters of the synapse and three parameters of the NetCon are fitted on two traces of synaptic currents following the MRF Tutorial.

The resulting values were rounded and thereby tested on criticality.
The MRF gives a delay of 0.059606 ms.
That could be rounded to 0.06ms or 0.1 ms and from biological point of view be set to 0 ms.

But any value except 0.06 (<=0.059 >=0.063) alters not only the time of onset but reduces the amplitude.
Additionaly the value of 1.06 ms (2.06, 3.06, ...)restores the old amplitude now shifted in time.

To control that behaviour, the same synapse was created outside the MRF
in a Test-Session.

[ted]

To evaluate this report, we will need some additional information.
First, what version of NEURON are you using (just copy the first line of the "welcome"
message that appears in NEURON's xterm), and on what platform (OS X, UNIX/Linux,
MSWin)?

Three parameters of the synapse and three parameters of the NetCon are fitted

The AMPA_S mechanism has 4 user-specifiable parameters: Cdur, Alpha, Beta,
and Erev. Which 3 do you mean? As far as the NetCon is concerned, this class has
only two user-specifiable params--weight and delay--so what "third parameter" was
the MRF adjusting?

The MRF is a complex tool, with many configurable parameters of its own. To help us
determine whether there is a bug, or an error of usage (which may require changes in
the MRF's user interface and/or documentation), it is essential to be able to reproduce
your observations. Please zip up the necessary source code (.hoc, .ses, .mod, the
MRF's .fd1 and .fd2 files, plus any data files such as psp waveforms) and email it to me
ted dot carnevale at yale dot edu

[ted]

As far as the NetCon is concerned, this class has
only two user-specifiable params--weight and delay--so what "third parameter" was
the MRF adjusting?

URig reminded me via email that the third parameter is threshold. My mistake. I promise
to never make another.