error when running a chain with varying nnu and meffsterile about cosmomc HOT 4 CLOSED

I guess you have stop_on_error=T? If False, that error should just reject the point without stopping. Do you have the parameters to reproduce the issue in CAMB?

from cosmomc.

Yes I have stop_on_error=T. If I set it to F, the following error occurs instead:
You need to normalize realistically to use lensing.
see http://cosmocoffee.info/viewtopic.php?t=94

The problem seems to be caused by small nnu (slightly larger than 3.046) and moderate meffsterile. One set of the problematic parameters related to neutrinos (other parameters are close to the standard values):
omnuh2 = 6.033281099611884E-003
nu_mass_degeneracies = 1.01533333333333 8.191413045527840E-002
nu_mass_fractions = 0.178640662473938 0.821359337526062
In CAMB, these parameters produce the error "HMCode INTEGRATE, Integration timed out".

from cosmomc.

Did you try the latest CAMB version, e.g. cmbant/CAMB@65b3f21 may be relevant? If that's not it, please could you give a full test case to reproduce it with CAMB, thanks.

from cosmomc.

Should be fixed in latest CAMB (let me know if you now see lots of warnings).

from cosmomc.