fix oscillation probability and add histograms for particle origin #40
Conversation
| @@ -223,7 +223,7 @@ void oscillate(grid_local_moment_type local_moments, grid_Gnu_type gnu) { | |||
| double p_survival = peq; | |||
| #else | |||
| double p_survival = | |||
| in_shallow ? peq : (1 - (1 - peq) * B / (A + SMALL)); | |||
| in_shallow ? peq : (1 - (1 - peq) * shallow / (deep + SMALL)); | |||
There was a problem hiding this comment.
Can you explain a bit on why you did ebar=total/stddev to compare it to gtemp? I am confused a bit on the dimensions of these quantities.
There was a problem hiding this comment.
In line 206,can you explain why you need the G!=0 condition if you are already checking for A!=0 and B!=0? G is dependent on the local angle (not talking about the global position angle) while A and B are not dependent on the local angle.. oh is it for ensuring not-oscillating where stddev was triggered as the comment seems to suggest?
There was a problem hiding this comment.
looks like you are not using the in_deep variable anywhere?
There was a problem hiding this comment.
Can you explain a bit on why you did ebar=total/stddev to compare it to gtemp? I am confused a bit on the dimensions of these quantities.
Actually, I think this is a bug. ebar is relative error. But g_temp is absolute size. So g_temp should be compared to stddev, not ebar.
In line 206,can you explain why you need the G!=0 condition if you are already checking for A!=0 and B!=0? G is dependent on the local angle (not talking about the global position angle) while A and B are not dependent on the local angle.. oh is it for ensuring not-oscillating where stddev was triggered as the comment seems to suggest?
Yes exactly. I set g = 0 when g < stddev, so it's used as a trigger.
looks like you are not using the in_deep variable anywhere?
Good catch.
There was a problem hiding this comment.
@Yurlungur , okay makes sense about the G != 0 and removal of the unused variable. Regarding the g_temp vs stddev vs ebar: Yes, indeed your fix was what I had in mind while commenting. I can't find any other obvious errors so will approve this. But to double check, g_temp you calculate by taking the distribution function f, and using the formula of the paper.. and stddev is calculated using quantities like w, N, which are whole numbers, where N is calculated by counting the number of superphotons in a given location and local angle bin.. so I would just make sure that those two quantities (gtemp and stddev) are reflecting the same things (which I think it is now)
There was a problem hiding this comment.
To double check, g_temp you calculate by taking the distribution function f, and using the formula of the paper.. and stddev is calculated using quantities like w, N, which are whole numbers, where N is calculated by counting the number of superphotons in a given location and local angle bin..
Correct.
| @@ -131,6 +131,7 @@ | |||
| bhl.config.set_cparm('RAD_NUM_TYPES', 4) | |||
| bhl.config.set_cparm('NEUTRINO_OSCILLATIONS', True) | |||
| bhl.config.set_cparm('FORCE_EQUIPARTITION', False) | |||
| bhl.config.set_cparm("RZ_HISTOGRAMS", True) | |||
There was a problem hiding this comment.
Just to confirm again, these RZ histograms will help to plot the R and Z origin locations of the photons in the bin under investigation, correct? so that we can track where the neutrinos in the bin came from and establish the effect of opacities/emissivities?
|
Thanks for the comments @payelmuk150 . Should be addressed now. Can you please re-review and approve if it looks good to you? |
I approve |
In the files changed tab, click on review at the top right and then click approve. |
Ah I see. Approve button missing... Will try to give you access. This is good enough though thanks. |
PR Summary
PR Checklist