-
I am using Cobaya with a custom likelihood and I am trying to sample it using a MCMC with MPI.
My issue is that each process samples the same points of the likelihood in the same order, i.e. the ou…
-
Hi
I am trying to run a simple chain with the native 2018 lensing and CMBMarg only lensing likelihood
planck_2018_lensing.clik:
path: ./external_modules/
clik_file: pl…
-
We've discussed this a bit before, but it continues to annoy me there's no easy way to reproduce things like "cobaya-run -r script.yaml" from a script, without loading and modifying the yaml.
Any …
-
I don't really understand what is being done here. I think, it would be helpful to add a little more description or some reference in the comment line.
https://github.com/sibirrer/hierArc/blob/b477…
-
I've been trying to run Planck-only chains for simple extensions to LCDM, and I've encountered random crashes that seem to be related to clik. The error message is not very useful, it only says:
```
…
-
Hello.
With cobaya 3.0.2 (but with 3.0.3 happens exactly the same) I tried to resume a run with custom likelihoods located in separate files. I am sure nothing has changed at all, though I got the fo…
-
Hi Antony and Jesus,
Currently it seems Cobaya returns get_fsigma8 and get_sigma_R, such that the redshift-dependent growth rate can be obtained from their ratio. However, as the growth rate is als…
-
Hello. With cobaya 3.0.3 I started running with CLASS and Planck18+H0 measurements, using `H0` instead of `theta_s` as input parameter. I've noticed that H0 gets into a very fast parameter block, whic…
-
I cannot resume chains that use an external theory code using the python_path argument. It seems like the chains do not store the python_path when saving
[input] {} (old) vs {‘python_path’: ‘/my/pa…
-
Hi,
I am using classy as the Boltzmann code and if i understand correctly when I launch:
$ cobaya-run input.yaml
each evaluation by classy is done on one thread only. While when I call (exter…