./GetDist data into Matlab Plotting - false/partial contour maps of likelihoods

Use of Cobaya. camb, CLASS, cosmomc, compilers, etc.
Post Reply
Luke Hart
Posts: 70
Joined: July 13 2015
Affiliation: University of Manchester

./GetDist data into Matlab Plotting - false/partial contour

Post by Luke Hart » July 16 2015

Hi,

So I'm incredibly sorry if this issue has arisen before but I checked the topic streams and found nothing so here goes:

I have used the ./getdist program as well, however my supervisor has told me to change the outputs of the created to .m files (since in our line MATLAB is easier to use and I haven't the foggiest when it comes to Python!)

I've got output scripts for the cmbalone and the 2D plots. Unfortunately, those the Gaussian 1D plots look acceptable, the contour plots (2D) are giving a false colour map and crazy contour lines

Could anyone explain to me why this is happening? The scripts run fine (though the 3d plot one doesnt) and the code is very dense with no error runs.

If it helps anyone to answer, I've used a MPI computer to run this and had to move to a local directory (given that the supercomputer doesn't have matlab) and I've only moved the output directory and the plot_data directory.

cmbalone_2D.pdf

Thanks in advance
Luke

Antony Lewis
Posts: 1941
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: ./GetDist data into Matlab Plotting - false/partial cont

Post by Antony Lewis » July 16 2015

You are plotting totally correlated variables (e.g. [tex]\Omega_m = 1-\Omega_\Lambda[/tex]). Probably other more useful plots will be OK, though matlab plots were never great. It would be a very good idea to switch to Python sooner rather than later.

Luke Hart
Posts: 70
Joined: July 13 2015
Affiliation: University of Manchester

./GetDist data into Matlab Plotting - false/partial contour

Post by Luke Hart » July 17 2015

Yeah these were preliminary graphs and I noticed the [tex] \Omega_m[/tex] vs. [tex]\Omega_\Lambda [/tex] graph and knew that already from our cosmology courses (well the assumption).

The thing that troubled me was the lack of colour present on the 2D graphs, but I've just gone through them again and they've come out quite nicely on MATLAB.

That being said, I've never used Python before so I'm gonna get equipped with it so I can graphically analyse the results through .py as well.

Thanks for your help Antony

Luke

Post Reply