CosmoMC: does not generate .converge file

Use of Cobaya. camb, CLASS, cosmomc, compilers, etc.
Post Reply
Ira *Wolfson
Posts: 69
Joined: January 24 2013
Affiliation: MPA
Contact:

CosmoMC: does not generate .converge file

Post by Ira *Wolfson » May 03 2017

Hi,

For some reason my CosmoMC runs do not generate the .converge file with convergence statistics.

CosmoMC runs fine, and checking the chains with GetDist mid-run generates graphs that approximate the known results nicely.

However, I use the .converge file to asses how much time is left for the run, and plan workload ahead....


My .ini file:

Code: Select all

#general settings
#Bicep-Keck-Planck, varying cosmological parameters
DEFAULT(batch2/BKPlanck.ini)

#Planck 2015, default just include native likelihoods (others require clik)
#DEFAULT(batch2/plik_dx11dr2_HM_v18_TT.ini)
#DEFAULT(batch2/lowTEB.ini)
#DEFAULT(batch2/lowl.ini)
DEFAULT(batch2/lensing.ini)

#Other Likelihoods
#DEFAULT(batch2/BAO.ini)
#DEFAULT(batch2/WiggleZ_MPK.ini)
#DEFAULT(batch2/MPK.ini)
#DEFAULT(batch2/WL.ini)

#general settings
DEFAULT(batch2/common.ini)

#e.g. to vary r in addition to standard 6:
#(for r>0 also need compute_tensors=T)
compute_tensors = T
param[r] = 0.03 0 2 0.04 0.04

redo_likelihoods = T
redo_theory =T
redo_cls = T
redo_pk = T
redo_skip = 0
redo_outroot = 
redo_thin = 1
redo_add = F
redo_from_text = F

#high for new runs
MPI_Max_R_ProposeUpdate = 30

propose_matrix= planck_covmats/base_TT_lowTEB_plik.covmat

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

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

num_threads = 8

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

start_at_bestfit =F
feedback=2
use_fast_slow = T

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

#Uncomment this if you don't want one 0.06eV neutrino by default
#num_massive_neutrinos=
#to vary parameters set param[name]= center, min, max, start width, propose width
#param[mnu] = 0 0 0 0 0


#for PICO install from https://pypi.python.org/pypi/pypico and download data file
#cosmology_calculator=PICO
#pico_datafile = pico3_tailmonty_v34.dat
#pico_verbose=F

rand_seed = 




Antony Lewis
Posts: 1936
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: CosmoMC: does not generate .converge file

Post by Antony Lewis » May 03 2017

CosmoMC generates a .converge_stat if run with MPI (after a while and burn in complete). .converge files are produced by getdist.

Ira *Wolfson
Posts: 69
Joined: January 24 2013
Affiliation: MPA
Contact:

CosmoMC: does not generate .converge file

Post by Ira *Wolfson » May 03 2017

I have

Code: Select all

burn_in = 0
, running with MPI...

Still no .converge_stat file generated.

I am running Cosmomc from the master branch downloaded on April 30th.

Jason Dossett
Posts: 97
Joined: March 19 2010
Affiliation: The University of Texas at Dallas
Contact:

CosmoMC: does not generate .converge file

Post by Jason Dossett » May 03 2017

With burn_in = 0, when running in MPI CosmoMC calculates its own burn in period (every parameter you are varying has to have changed at least 51 times).

Check the outputs from stdout, you should see something like
Chain 1 MPI done ''burn'', Samples = xxxx, like = xxxx
Time: xxxx output lines= xxxx
If you don't see that, your job has not completed burn in.

Another way to check and see if your chains have finished burn in is to check if you have any .chk files if checkpointing is enabled.

Post Reply