Number of degrees of freedom in CosmoMC runs
-
- Posts: 3
- Joined: March 28 2013
- Affiliation: Swinburne University of Technology
- Contact:
Number of degrees of freedom in CosmoMC runs
Hi all,
My apologies, if this is trivial to figure out, but I failed to find the number degrees of freedom used in my cosmoMc runs. I also could not find mention in the Planck XVI paper.
Is there any file that states the dof specifically (or number of data points, and parameters used so I could infer myself)?
I am currently running the October 2013 version. I am using the vanilla parameters (flat LCDM), with Planck+WP, without ACT or SPT.
If it is of importance to answer the question I use in the ini file:
DEFAULT(batch1/CAMspec_defaults.ini)
DEFAULT(batch1/lowl.ini)
DEFAULT(batch1/lowLike.ini)
DEFAULT(batch1/common_batch1.ini)
(apologies if I am missing any critical information, I am a new-be to cosmomc)
My apologies, if this is trivial to figure out, but I failed to find the number degrees of freedom used in my cosmoMc runs. I also could not find mention in the Planck XVI paper.
Is there any file that states the dof specifically (or number of data points, and parameters used so I could infer myself)?
I am currently running the October 2013 version. I am using the vanilla parameters (flat LCDM), with Planck+WP, without ACT or SPT.
If it is of importance to answer the question I use in the ini file:
DEFAULT(batch1/CAMspec_defaults.ini)
DEFAULT(batch1/lowl.ini)
DEFAULT(batch1/lowLike.ini)
DEFAULT(batch1/common_batch1.ini)
(apologies if I am missing any critical information, I am a new-be to cosmomc)
-
- Posts: 1943
- Joined: September 23 2004
- Affiliation: University of Sussex
- Contact:
Re: Number of degrees of freedom in CosmoMC runs
The Planck likelihoods are not all normalized to match a chi-square value (e.g. low-L is actually negative), so I'm not sure it's a very well defined question/answer.
-
- Posts: 3
- Joined: March 28 2013
- Affiliation: Swinburne University of Technology
- Contact:
Number of degrees of freedom in CosmoMC runs
Would it make sense, then, to do AIC (Akaike information criterion) and BIC (Bayesian information criterion) comparisons between results when using different data sets and adding parameters?
For BIC, e.g, I would need the number of data points used.
Where can I find the number of data points used for fitting Planck+WP including low-L?
For BIC, e.g, I would need the number of data points used.
Where can I find the number of data points used for fitting Planck+WP including low-L?
-
- Posts: 8
- Joined: November 28 2015
- Affiliation: National Taiwan University
Number of degrees of freedom in CosmoMC runs
I encounter the same problem to find the degrees of freedom when using Planck data set. I tried to find the number of data points, but fail to find it in papers released by Planck.