CosmoMC - new parameters not converging

[vivian sabla]

Hello!

I have modified CAMB and CosmoMC [version: Oct. 2018] to include a new parameter and everything seems to be working fine. From running camb for different parameter values I can see by inspection that the new parameter should be constrained but when I run CosmoMC the step-size in the chains changes to be too large resulting in the chains never converging. This happen no matter what I set the center, min, max or widths to.

Is there a way to turn off dynamical step size in CosmoMC? Or is there something I am missing about how to implement new parameters into CosmoMC?

Changes I have made to CosmoMC:
Calculator_CAMB.f90: added translation from CAMB to CosmoMC parameter (i.e. P%variable = CMB%variable)
CosmologyParametrizations.f90: increased number of hard parameters in SetTheoryParameterNumbers in TP_Init and added the new parameter in SetforH to the Params array
CosmologyTypes.f90: added definitions of new parameter to CMBParams
params_CMB.paramnames: added the new parameter

Any help would be very much appreciated!

[Antony Lewis]

You could run with action=4 and set test_output_root to check the outputs from cosmomc are as expected from the parameter cetre values.

[vivian sabla]

Thanks! I have tried that and the output Cl's from CosmoMC are the same no matter what center values I use. When running camb on its own, the values for my new parameters do shift the Cl spectrum as expected however when camb is run through CosmoMC it seems not to make a difference.

The edits I have made to CosmoMC are in my first post. Any idea what I am doing wrong?

Thanks!