[MontePython+hi_class] compilation issue

Use of Cobaya. camb, CLASS, cosmomc, compilers, etc.
Post Reply
TUMURTUSHAA GANSUKH
Posts: 6
Joined: June 22 2020
Affiliation: Jeju National University

[MontePython+hi_class] compilation issue

Post by TUMURTUSHAA GANSUKH » July 14 2020

Hi?

I have some compiling issues with MontePython (+ hi_class) whenever I try to use either Planck or Supernova (both JLA or Pantheon) data. Both MontePython and hi_class have "successfully" installed on my macOS. I am saying successful because I have no problem with running and reproducing exercises provided on their respective webpage if I use BAO data or some given data (and likelihoods) that comes with MontePython installation.
The problem occurs only when using Planck or JLA (and Pantheon) data.

Without further due, let me quickly lead you to how I compile into something I could not figure the way out.

First, in the MontePython directory, my "default.conf" file looks like:

Code: Select all

root = '/Users/kosmos'

#path['cosmo']		= root+'/Documents/class_public/'

path['cosmo']		= root+'/hi_class_public/'
path['clik']            = root+'/montepython_public_3.2dev_Python3/planck/code/plc_3.0/plc-3.01/'
path['JLA']		   = root+'/montepython_public_3.2dev_Python3/data/JLA/'
path['Pantheon']  = root+'/montepython_public_3.2dev_Python3/data/Pantheon/'
and you can see where both Planck likelihood and JLA data are installed and stored. There was no problem during installation of plc-3.01 and "source"-ing "clik_profile.sh" was successful. The JLA directory contains the following files:

Code: Select all

montepython_public_3.2dev_Python3 kosmos$ ls data/JLA/
JLA.paramnames		jla.dataset		jla_mub.txt~		jla_simple.dataset~	jla_v0b_covmatrix.dat	jla_vb_covmatrix.dat
LICENSE.txt		jla_lcparams.txt	jla_mub_covmatrix.dat	jla_v0_covmatrix.dat	jla_va_covmatrix.dat	makefile
ReadMe.txt		jla_mub.txt		jla_simple.dataset	jla_v0a_covmatrix.dat	jla_vab_covmatrix.dat
montepython_public_3.2dev_Python3 kosmos$
and "my_JLA.param" file is shown as follows:

Code: Select all

data.experiments=['JLA']

data.parameters['Omega_cdm'] 	     = [0.2562, None,   None,    0.008,    1,       'cosmo']

data.parameters['alpha'] 	         = [0.15, None, None, 0.001, 1, 'nuisance']
data.parameters['beta'] 	         = [3.559, None, None, 0.02, 1, 'nuisance']
data.parameters['M'] 	             = [-19.02, None, None, 0.004, 1, 'nuisance']
data.parameters['Delta_M'] 	     = [-0.10, None, None, 0.004, 1, 'nuisance']

data.parameters['Omega_m']  = [0,     -1,  -1,   0,           1,      'derived']

data.cosmo_arguments['Omega_b'] 	     = 0.048
data.cosmo_arguments['h']         	             = 0.68
data.cosmo_arguments['T_cmb']              = 2.726
data.cosmo_arguments['N_eff']                = 3.046
data.cosmo_arguments['N_ncdm']              = 0

data.N=10
data.write_step=5
With this setup, now I can do:

Code: Select all

 
$ mpirun -np 4 python montepython/MontePython.py run -p input/my_JLA.param -o my_JLA/ -N 1000 
which gives me the message as follows

Code: Select all

Running Monte Python v3.2.0

with CLASS v2.7.2
 /!\ Detecting empty folder, logging the parameter file

Testing likelihoods for:
 ->
JLA

Traceback (most recent call last):
  File "montepython/MontePython.py", line 38, in <module>
    sys.exit(mpi_run())
  File "/Users/kosmos/montepython_public_3.2dev_Python3/montepython/run.py", line 101, in mpi_run
    custom_command, comm, nprocs)
  File "/Users/kosmos/montepython_public_3.2dev_Python3/montepython/run.py", line 191, in safe_initialisation
    cosmo, data, command_line, success = initialise(custom_command)
  File "/Users/kosmos/montepython_public_3.2dev_Python3/montepython/initialise.py", line 66, in initialise
    data = Data(command_line, path)
  File "/Users/kosmos/montepython_public_3.2dev_Python3/montepython/data.py", line 363, in __init__
    self.initialise_likelihoods(self.experiments)
  File "/Users/kosmos/montepython_public_3.2dev_Python3/montepython/data.py", line 492, in initialise_likelihoods
    elem, elem, folder, elem))
  File "<string>", line 1, in <module>
  File "/Users/kosmos/montepython_public_3.2dev_Python3/montepython/likelihoods/JLA/__init__.py", line 64, in __init__
    self.C00 = self.read_matrix(self.mag_covmat_file)
  File "/Users/kosmos/montepython_public_3.2dev_Python3/montepython/likelihood_class.py", line 2472, in read_matrix
    matrix = read_table(path).as_matrix().reshape((length, length))
  File "/opt/anaconda3/lib/python3.7/site-packages/pandas/core/generic.py", line 5274, in __getattr__
    return object.__getattribute__(self, name)
AttributeError: 'DataFrame' object has no attribute 'as_matrix'

and stuck there. I do not know what I have been doing wrong or how to figure out the way to resolve the issue.
I somehow feel (deep down in my hearth) that this is not a serious issue. Therefore, a solution might be simple but I do not get it.

Can anyone help me with this? Thank you in advance.

Post Reply