DES : Execution fails on C_l's negative and message TTanhReionization_zreFromOptDepth: Did not converge to optical depth

Use of Cobaya. camb, CLASS, cosmomc, compilers, etc.
Post Reply
Dournac Fabien
Posts: 74
Joined: May 18 2019
Affiliation: IRAP
Contact:

DES : Execution fails on C_l's negative and message TTanhReionization_zreFromOptDepth: Did not converge to optical depth

Post by Dournac Fabien » March 01 2022

Hi !

I tried to make run chains with DES survey but cosmomc failed up to 3 hours of computing.

Here the des_curvature.ini that I am using :

Code: Select all

#general settings, test file without require Planck likelihood code to be installed.
#see test_planck.ini for test file including Planck clik likelihoods

#DES 1-yr joint
DEFAULT(batch3/DES_lensing.ini)

#general settings
DEFAULT(batch3/common.ini)

param[omegak] = 0. -0.5 0.5 0.01 0.01

#high for new runs, to start learning new proposal matrix quickly
MPI_Max_R_ProposeUpdate = 30

propose_matrix= planck_covmats/base_TT_lensing_lowE_lowl_plikHM.covmat

#Folder where files (chains, checkpoints, etc.) are stored
root_dir = chains/

#Root name for files produced
file_root=des_omegak
#action= 0 runs chains, 1 importance samples, 2 minimizes
#use action=4 just to quickly test likelihoods
action = 0

#Don't need to set this if OMP_NUM_THREADS set appropriately
num_threads = 0

#if you want to get theory cl for test point
#test_output_root = output_cl_root

start_at_bestfit =F
feedback=1
use_fast_slow = T

#turn on checkpoint for real runs where you want to be able to continue them
checkpoint = T

#sampling_method=7 is a new fast-slow scheme good for Planck
sampling_method = 7
dragging_steps  = 3
propose_scale = 2

#Set >0 to make data files for importance sampling
indep_sample=10

#these are just small speedups for testing
get_sigma8=T
1) You can see that I have put a prior on Omega_k, I don't know if it is useful : if someone could confirm the utility of adding this prior.

2) Secondly, below the log file of the run which fails :

Code: Select all

Chain:0 drag accpt:  0.4849885     fast/slow   58.51421     slow:        2392
 Chain:1 drag accpt:  0.5351521     fast/slow   58.67863     slow:        1867
 Chain:3 drag accpt:  0.3510825     fast/slow   58.79749     slow:        2953
 Chain:2 drag accpt:  0.4091374     fast/slow   58.76303     slow:        2819
 Chain:0 drag accpt:  0.4835083     fast/slow   58.42980     slow:        2443
 Chain:3 drag accpt:  0.3547736     fast/slow   58.73333     slow:        3000
 Chain:0 drag accpt:  0.4845815     fast/slow   58.36338     slow:        2496
 Chain:1 drag accpt:  0.5370844     fast/slow   58.58159     slow:        1912
 Chain           3  MPI communicating
 Chain           2  MPI communicating
 Chain           4  MPI communicating
 Chain           1  MPI communicating
 Current convergence R-1 =    8.874245      chain steps =        2777
 slow changes        2534  power changes           0
 updating proposal density
 Chain:3 drag accpt:  0.3547297     fast/slow   58.65588     slow:        3060
 Chain:0 drag accpt:  0.4836976     fast/slow   58.29717     slow:        2544
 Chain:2 drag accpt:  0.4063429     fast/slow   58.61038     slow:        2890
 Chain:1 drag accpt:  0.5373135     fast/slow   58.40754     slow:        1963
 Chain:0 drag accpt:  0.4808362     fast/slow   58.16468     slow:        2605
 TTanhReionization_zreFromOptDepth: Did not converge to optical depth
 tau =  0.211094049991010      optical_depth =   0.305066709379731
   50.0000000000000        49.9938964843750
 (If running a chain, have you put a constraint on tau?)
Chain:3 drag accpt:  0.3572417     fast/slow   58.57617     slow:        3105
 Chain:1 drag accpt:  0.5339105     fast/slow   58.21110     slow:        2018
 Chain:0 drag accpt:  0.4655002     fast/slow   58.03076     slow:        2698
 Chain:3 drag accpt:  0.3607543     fast/slow   58.46537     slow:        3148
 Chain:2 drag accpt:  0.4002541     fast/slow   58.47289     slow:        2969
 Chain:1 drag accpt:  0.5347092     fast/slow   58.10993     slow:        2065
 Chain:0 drag accpt:  0.4664723     fast/slow   57.93632     slow:        2748
 Calculator_CAMB: negative C_l (could edit to silent error here)
 MpiStop:            1
Abort(-1644241136) on node 1 (rank 1 in comm 0): application called MPI_Abort(MPI_COMM_WORLD, -1644241136) - process 1
2.1) I don't understand why I get the message :

Code: Select all

TTanhReionization_zreFromOptDepth: Did not converge to optical depth
 tau =  0.211094049991010      optical_depth =   0.305066709379731
 
Have got to put a prior on tau ? I tried also to make run by putting :

Code: Select all

param[omegak] = 0. -0.2 0.2 0.01 0.01
param[tau] = 0.05430842
or even only with :

Code: Select all

param[tau_reio] = 0.05430842 0.04 0.06 0.001 0.001 
but it gives very strange cosmology contours. Which is the correct syntax to avoid this message
TTanhReionization_zreFromOptDepth: Did not converge to optical depth
?

2.2) You can notice that program stops since :

Code: Select all

 Calculator_CAMB: negative C_l (could edit to silent error here)
 MpiStop:            1
From what I understood, he stops because it has evaluted negative C_l, isn't it ?

Any help is welcome to solve the points 1), 2.1) and 2.2)

Best regards

Dournac Fabien
Posts: 74
Joined: May 18 2019
Affiliation: IRAP
Contact:

Re: DES : Execution fails on C_l's negative and message TTanhReionization_zreFromOptDepth: Did not converge to optical d

Post by Dournac Fabien » March 08 2022

It seems that prior on Omega_k that I have put is too "large" (Surely CAMB produces surely negative [math] in this case)

Instead of :

Code: Select all

param[omegak] = 0. -0.5 0.5 0.01 0.01
I have set :

Code: Select all

param[omegak] = 0. -0.1 0.1 0.01 0.01
and for the moment, the current execution doesn't give any more error messages of kind :
Execution fails on C_l's negative and message TTanhReionization_zreFromOptDepth: Did not converge to optical depth
1) Now, is it correct if I put a such strong constraint on Omega_k ?

2) the last fiveth value (0.01) in

Code: Select all

param[omegak] = 0. -0.1 0.1 0.01 0.01 
must be lower than the value just before (fourth value), mustn't it ? since it is the step adopted for the chain of Omega_k, right ?

Regards

Post Reply