CAMB: question on extra evolution variable
-
- Posts: 24
- Joined: January 11 2019
- Affiliation: University of Michigan
CAMB: question on extra evolution variable
Hi Anthony,
I added a set of extra evolution variables for dark radiation, and I have some concerns about my practice here.
What I did was to add EV%dr_ix as the equation index of dark radiation, and add the same number of equations as massless neutrino in subroutine 'SetupScalarArrayIndices', then did adjustments to subroutine 'initial' and derivs. If my understanding is right, now when I'm doing GaugeInterface_EvolveScal(EV,tau,y,State%Transfer_Times(itf),tol1,ind,c,w), y variable has clx,q,pi for all components including my extension for dr, but since I did nothing to EV%OutputTransfer, I won't get the transfer data of dr in the end.
My question is, will this practice mess up the output of other transfer functions? Is not getting dr perturbation evolution the only aftermath that I didn't do OutputTransfer and only adapting 'y' structure in the code? Specifically, I was worried about that EV%OutputTransfer components mismatching the differential equation components would cause some transfer data to be put into wrong boxes.
Thanks,
Angela
I added a set of extra evolution variables for dark radiation, and I have some concerns about my practice here.
What I did was to add EV%dr_ix as the equation index of dark radiation, and add the same number of equations as massless neutrino in subroutine 'SetupScalarArrayIndices', then did adjustments to subroutine 'initial' and derivs. If my understanding is right, now when I'm doing GaugeInterface_EvolveScal(EV,tau,y,State%Transfer_Times(itf),tol1,ind,c,w), y variable has clx,q,pi for all components including my extension for dr, but since I did nothing to EV%OutputTransfer, I won't get the transfer data of dr in the end.
My question is, will this practice mess up the output of other transfer functions? Is not getting dr perturbation evolution the only aftermath that I didn't do OutputTransfer and only adapting 'y' structure in the code? Specifically, I was worried about that EV%OutputTransfer components mismatching the differential equation components would cause some transfer data to be put into wrong boxes.
Thanks,
Angela
-
- Posts: 1944
- Joined: September 23 2004
- Affiliation: University of Sussex
- Contact:
Re: CAMB: question on extra evolution variable
I think not setting OutputTransfer should be fine if you don't want to look at/calculate anything with the transfer functions for your new components.
-
- Posts: 24
- Joined: January 11 2019
- Affiliation: University of Michigan
Re: CAMB: question on extra evolution variable
Thanks Antony, sorry for typing your name wrong! Another further concern is, would lacking this part of dr transfer function information cause error in the integration of Cls? and later in the calculation of lensed Cl?Antony Lewis wrote: ↑June 13 2019I think not setting OutputTransfer should be fine if you don't want to look at/calculate anything with the transfer functions for your new components.
-
- Posts: 1944
- Joined: September 23 2004
- Affiliation: University of Sussex
- Contact:
Re: CAMB: question on extra evolution variable
No, the sources for the CL are calculated as separate things.
-
- Posts: 24
- Joined: January 11 2019
- Affiliation: University of Michigan
Re: CAMB: question on extra evolution variable
Ok, thanks!