Some troubles about adding new parameters in CosmoMC

Use of Cobaya. camb, CLASS, cosmomc, compilers, etc.
Post Reply
Yutong Wang
Posts: 9
Joined: May 06 2014
Affiliation: UCAS

Some troubles about adding new parameters in CosmoMC

Post by Yutong Wang » October 01 2016

Hi everyone!

I am trying to change cosmomc for adding two extra, new parameters.
In new model, the scalar power spectrum multiplied by an exponential cut-off:
P_s(k)=P_0(k)*(1-exp(-(k/k_c)^lc))),

I wanna to get the bestfit value of "k_c" and "lc", following the procedure in CosmoMC readme I added the "k_c" successfully.

for "lc", I use an easier and faster way, to modify the parameter "n_run" which not be use in my model and use it to instead of "lc".

However, For Planck TT,TE,EE+lowP, the best-fit model has n_run=0.63, and ln(k_c/Mpc^-1)=-7.96. (the correct bestfit is lc=0.53(n_run=0.53) and ln(k_c/Mpc^-1)=-7.98) Maybe there is something wrong with the modification about n_run.

I made some modifications for "n_run" in the following files:
in cosmomc/CAMB/power_tilt.f90, I modify the scalar power-law spectrum
as my new model.
in cosmomc/source/CosmologyParameterizations.f90, I modify the

Code: Select all

derived(ix) = CMB%InitPower(ns_index) +CMB%InitPower(nrun_index)*lograt +&
            CMB%InitPower(nrunrun_index)*lograt**2/2

as

Code: Select all

derived(ix) = CMB%InitPower(ns_index) +&
            CMB%InitPower(nrunrun_index)*lograt**2/2
in cosmomc/batch2/params_CMB_defaults.ini, I make the modification

Code: Select all

param[nrun] = 0.6, 0, 1, 0.01, 0.01
and in others files, I didn't make modification about n_run, and anyone can help me to point out the mistake in this process ? It puzzled me for a long time, thanks a lot!

PS:the version of CosmoMC is 2015 July, and ifort is 14+

Post Reply