pyNLO/PyNLO

Disabling Raman while leaving self-steepening breaks things

DanHickstein opened this issue · 7 comments

DanHickstein commented

So, in the Dudley example, either of these options work fine:

evol = SSFM.SSFM(dz = dz, local_error = 0.001, USE_SIMPLE_RAMAN = True)
evol = SSFM.SSFM(dz = dz, local_error = 0.001, USE_SIMPLE_RAMAN = True,  disable_Raman=True, disable_self_steepening=True)

However,

evol = SSFM.SSFM(dz = dz, local_error = 0.001, USE_SIMPLE_RAMAN = True,  disable_Raman=True)

has overflow warnings and the pulse energy drops to zero.

Any idea what is going on?

ycasg commented

What happens when you set disable_Raman=True and give no other
Raman-related parameters? The disable Raman flag should supercede simple
Raman (so if that's working, it the no Raman case is broken. Which is very
possible.)

On Fri, Mar 18, 2016 at 2:49 PM, Danhickstein notifications@github.com
wrote:

So, in the Dudley example, either of these options work fine:

evol = SSFM.SSFM(dz = dz, local_error = 0.001, USE_SIMPLE_RAMAN = True)
evol = SSFM.SSFM(dz = dz, local_error = 0.001, USE_SIMPLE_RAMAN = True, disable_Raman=True, disable_self_steepening=True)

However,

evol = SSFM.SSFM(dz = dz, local_error = 0.001, USE_SIMPLE_RAMAN = True, disable_Raman=True)

has overflow warnings and the pulse energy drops to zero.

Any idea what is going on?

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub
#20

Gabriel Ycas, PhD

DanHickstein commented

The USE_SIMPLE_RAMAN doesn't seem to change the behavior when Raman is disabled, as you suspected.

ycasg commented

OK, then the problem is around the nonlinear step. As inherited from
LaserFOAM, the no-Raman case is completely different from the Raman-enabled
case. This is not the best thing, and the solution might be setting the
Raman gain to zero when the disable flag is passed.

On Fri, Mar 18, 2016 at 4:03 PM, Danhickstein notifications@github.com
wrote:

The USE_SIMPLE_RAMAN doesn't seem to change the behavior when Raman is
disabled, as you suspected.

—
You are receiving this because you commented.
Reply to this email directly or view it on GitHub
#20 (comment)

Gabriel Ycas, PhD

DanHickstein commented

Yeah, I like the idea of just setting the Raman gain to zero. Howe exactly does one do that? Does that mean that the tau values should be zero? or the f_R?

ycasg commented

I think you can just go one step upstream and in CalculateRamanResponseFT,
add an initial

if self.disable_raman:
self.R[:]=0
else:
...existing code...

On Fri, Mar 18, 2016 at 4:13 PM, Danhickstein notifications@github.com
wrote:

Yeah, I like the idea of just setting the Raman gain to zero. Howe exactly
does one do that? Does that mean that the tau values should be zero? or the
f_R?

—
You are receiving this because you commented.
Reply to this email directly or view it on GitHub
#20 (comment)

Gabriel Ycas, PhD

DanHickstein commented

I like this idea.

self.R[:] = 0 stops all modification of the input pulse under any circumstances.

But, I think that self.R[:] = 1 does the trick.

Does that seem reasonable?

DanHickstein commented

#21 seems to have addressed this to my satisfaction. Closing this issue.