error when running ./cosmomc test.ini

Use of Healpix, camb, CLASS, cosmomc, compilers, etc.
Post Reply
Nishanth Sasankan
Posts: 27
Joined: September 07 2016
Affiliation: university of notre dame

error when running ./cosmomc test.ini

Post by Nishanth Sasankan » December 02 2017

is anyone familiar with this error

Time for theory: 8.89974
Time for smica_g30_ftl_full_pp: 9.000301361083984E−004
Time for BKPlanck_detset_comb_dust: 7.591247558593750E−004
loglike chi-sq
22.333 44.665 CMB: BKPLANCK = BKPlanck_detset_comb_dust
5.413 10.826 CMB: lensing = smica_g30_ftl_full_pp
Test likelihoods done, total logLike, chi-eq = 27.746 55.491
Expected likelihoods, total logLike, chi-eq = 28.337 56.674
** Likelihoods do not match **
MpiStop: 0
application called MPI_Abort(MPI_COMM_WORLD, 18299264) - process 0
[unset]: write_line error; fd=−1 buf=:cmd=abort exitcode=18299264
:
system msg for write_line failure : Bad file descriptor

I just wanted to know what is causing the MPI_Abort
Thanks
Nishanth

Antony Lewis
Posts: 1353
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: error when running ./cosmomc test.ini

Post by Antony Lewis » December 02 2017

Probably just because the test failed.

Did you change some setting to make the test fail?

Nishanth Sasankan
Posts: 27
Joined: September 07 2016
Affiliation: university of notre dame

error when running ./cosmomc test.ini

Post by Nishanth Sasankan » December 02 2017

All I did was write a subroutine to generate an array of appf and wppf depending on wa_ppf in equations_ppf.f90. But this shouldn't be any different from reading (a,w) from file in camb. Also I did add a range for wa in params_CMB_defaults.ini in batch2 folder.

I checked running my camb to see if it had any errors, but it worked fine.

Thanks
Nishanth

Antony Lewis
Posts: 1353
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: error when running ./cosmomc test.ini

Post by Antony Lewis » December 03 2017

If the default install runs fine, and you are testing with w=−1, wa=0, then you probably have a bug. If you've changed the default model, then of course you can expect the test comparison to LCDM to fail.

Nishanth Sasankan
Posts: 27
Joined: September 07 2016
Affiliation: university of notre dame

error when running ./cosmomc test.ini

Post by Nishanth Sasankan » December 03 2017

I changed the code to not have LCDM. So that is probably why it was happening.

Post Reply