question about the oscillations example

[xyang]

I was trying to read the spatiotemporal accuracy vs. computational speed in
http://www.neuron.yale.edu/neuron/paper ... cetime.pdf
in page 9, line5 and 6, there is a example for showing to calculate the oscillations, what I calculated is delta(t) <= 0.32ms. I always could not figure out what am I wrong. delta(x) is 20, lambda is 500, tau is 0.016.
Thanks for the help!
Xianghong

[ted]

Thanks for bringing this to my attention. You may have discovered an error in the preprint.
If DX^2 << 1, oscillations will occur if DT/DX^2 > 1/2 (actually 1/2 - epsilon, where epsilon is
very small compared to 1/2). In other words, the DX should be squared. I will check the
derivation of the criterion, and generate a corrected preprint if necessary. It will also be
necessary to add this to the list of errata for The NEURON Book (affects the text box on
page 70). I will also check the publised paper for this error and ask The Neuroscientist
to announce an erratum if appropriate.

[ted]

Indeed, the DX should be squared. I have already placed a revised version of the
manuscript on NEURON's WWW site--see the link to
spacetime_rev1.pdf
on
http://www.neuron.yale.edu/neuron/bib/nrnpubs.html
--and will relay an erratum report to The Neuroscientist.
Also, I will make a new entry in this forum's thread
"The NEURON Book: errata and discussion"
https://www.neuron.yale.edu/phpBB2/viewtopic.php?t=316