Cosmology context plotting : Covariance matrix - Script to plot Omega_m vs Omega_Lambda diagram in Python

I have covariance matrices and Markov chains coming from the SCP (Supernova Cosmology Project) and I would like to plot all these data in the particular diagram Omega_Lambda vs Omega_m or w vs Omega_m.

For example, I would like to get this kind of figures :

and

I would prefer to write a Python script.

Update 1

I managed to get the covariance matrice available on: http://supernova.lbl.gov/Union/figures/SCPUnion2.1_covmat_nosys.txt from the SCP (SuperNova Cosmology Project): http://supernova.lbl.gov/Union/

Once downloaded, I computed the associated Fisher matrix. But I realize that the size of the matrix is 307x307.

I don't know how to exploit this big and only one covariance/Fisher matrix.

On the figures above, there are 3 probes (CMB, SNIa, and BAO) and in grey, I guess the cross of the 3 previously cited probes.

How can I handle this large matrix? Maybe the authors have made a synthesis of the 3 probes + cross-correlation in a unique matrix. If this is the case, how to break down this only one matrix to plot individually each probe?

Update 2

I recall that Covariance matrix is equal to the inverse of Fisher matrix (matrix representing the information that we have on each cosmological parameter). Then, the covariance matrix gives variances on diagonal elements and covariances on off-diagonal elements.

Update 3

Below the Python script used with getdist tool (pip install getdist) with Fisher matrix coming the inverse of Covariance downloaded above :

# Show plots inline, and load main getdist plot module and samples class
from __future__ import print_function
import sys, os
sys.path.insert(0,os.path.realpath(os.path.join(os.getcwd(),'..')))
import numpy as np
import getdist
from getdist import plots
import matplotlib.pyplot as plt
import matplotlib as mpl
# The plotting scripts also let you plot Gaussian (or Gaussian mixture) contours
from getdist.gaussian_mixtures import GaussianND
from getdist.gaussian_mixtures import Mixture2D
print('GetDist Version: %s, Matplotlib version: %s'%(getdist.__version__, plt.matplotlib.__version__))

## Load Fisher matrix and invert it to get Covariance array
# Load Fisher matrix
# Invert to get Covariance matrix
COV_File1 = np.linalg.inv(File1)[2:4,2:4]

#  Example of pre-defined location for legend
# Mean of each cosmo parameters : respect order of generated "Big" Fisher matrix
mean = [0.28, 0.72]
# Names and labels
names = ['Omeega_m', 'Omega_L']
labels = [ r'$$\Omega_m$$','$$\Omega_{\Lambda}$$']
matrix1 = GaussianND(mean, COV_File1, labels = labels, names = names)
# Plot triplot
plt.rcParams['text.usetex'] = True
plt.rc_context({'axes.autolimit_mode': 'round_numbers'})
g = plots.get_subplot_plotter()
g.settings.figure_legend_frame = True
g.settings.legend_fontsize = 8
g.settings.axes_labelsize = 15
g.settings.axes_fontsize  = 10
g.settings.title_limit_fontsize = 18
g.settings.axis_tick_x_rotation = 45
g.settings.axis_tick_y_rotation = 45
g.settings.fig_width_inch = 5
g.settings.fig_width_inch = 5
g.settings.axis_tick_max_labels = 10
g.triangle_plot([matrix1],
names,
filled = True,
legend_labels = [ r'$$\Omega_{\Lambda}$$'+' vs '+r'$$\Omega_{m}$$'],
legend_loc = 'upper right',
contour_colors = ['blue'],
)
# Save triplot
g.export('output_OmegaL_Omegam.pdf')


Unfortunately with this current script, I get degenerated 2D joint distribution between Omega_Lambda and Omega_m :

• I have to question if this is the appropriate exchange for this question. You're plotting astronomy related concepts, but fundamentally your question is about plotting in python. It might be more appropriate to ask this on another exchange like Stack Overflow. Oct 14 at 17:25
• @zephyr . I know but people on stack Overflow told me that background was astrophysics and advised me to ask on another forum. So I try my chance here Oct 14 at 17:39