CosmoMC: Missing H0 in Getdist output

Use of Cobaya. camb, CLASS, cosmomc, compilers, etc.
Post Reply
Alan Zablocki
Posts: 8
Joined: April 03 2010
Affiliation: University of Chicago, KICP

CosmoMC: Missing H0 in Getdist output

Post by Alan Zablocki » February 08 2011

When I run getdist I get this error:

ParamNames: Unknown parameter name H0

When plotting the 1D marginalized distributions or confidence regions H0 is absent in the.ps file. The file_root.txt file has 22 columns, and am fairly certain the last column is H0. Files such as .paramnames, and .converge are missing H0. Incidentally, I am missing the .margestats and .likestats files! Any idea why that would happen?

Other than that, H0 is missing as a parameter from all other files. I am probably not setting something, however it would have been like this since downloading the software, any ideas on what might be going wrong here or what I have not set? I supply my params and distparams files.Thanks in advance.

params.ini

Code: Select all

file_root =chains/planck_chain4/planck4
action = 0
samples = 90000
feedback = 2
temperature = 1
cmb_numdatasets = 1
cmb_dataset1 = data/planck.dataset
nonlinear_pk = F
use_CMB = T
use_HST = F
use_mpk = F
use_BAO = F
use_clusters = F
use_BBN = F
use_Age_Tophat_Prior = T
use_SN = F
use_lya = F
use_min_zre = 0
data_dir = data/
get_sigma8 = F
sampling_method = 1
directional_grid_steps = 20
use_fast_slow = F
propose_matrix = params_CMB.covmat
estimate_propose_matrix = F
delta_loglike = 2
propose_scale = 2.4
oversample_fast = 1
indep_sample = 1
burn_in = 0
num_threads = 0
MPI_Converge_Stop = 0.03
MPI_StartSliceSampling  = F
MPI_Check_Limit_Converge = F
MPI_Limit_Converge = 0.025
MPI_Limit_Converge_Err = 0.2
MPI_Limit_Param = 0
MPI_LearnPropose = T
MPI_R_StopProposeUpdate = 0
MPI_Max_R_ProposeUpdate = 2
MPI_Max_R_ProposeUpdateNew = 30
rand_seed = 
checkpoint = F
compute_tensors = F
pivot_k = 0.05
inflation_consistency = F
bbn_consistency=T
CMB_lensing = F
#since the data you input, the Cl do not have CMB lensing on, just yet.
accuracy_level = 1
redo_likelihoods = T
redo_theory = F
redo_cls = F
redo_pk = F
redo_skip = 0
redo_outroot = 
redo_thin = 1
redo_add = F
redo_from_text = F
redo_likeoffset =  0
param[omegabh2] = 0.0223 0.005 0.1 0.001 0.001
param[omegadmh2] = 0.105 0.01 0.99 0.01 0.01
param[theta] = 1.04 0.5 10 0.002 0.002
param[tau] = 0.09 0.01 0.8 0.03 0.03
param[omegak] = 0 0 0 0 0
param[fnu] = 0 0 0 0 0
param[w] = -1 -1 -1 0 0
param[ns] = 0.95 0.5 1.5 0.02 0.01
param[nt] = 0 0 0 0 0
param[nrun] = 0 0 0 0 0
param[logA] = 3 2.7 4 0.01 0.01
param[r] = 0 0 0 0 0
param[asz]= 1 0 2 0.4 0.4
distparams.ini

Code: Select all

columnnum  = 0
file_root = chains/planck_chain4/planck4
out_root = 
out_dir = 
plot_data_dir = plot_data/
parameter_names = 
cool = 1
chain_num = 0
first_chain = 
exclude_chain = 
num_bins = 20
ignore_rows = 1
B&W = F
matlab_version = 10
no_plots = F
no_tests = F
auto_label = F
samples_are_chains = T
compare_num = 0
compare1 = basic6_cmb
plot_meanlikes = T
shade_meanlikes = T
thin_factor = 0
make_single_samples = F
single_thin = 4
adjust_priors = F
map_params = F
smoothing = T
num_contours = 2
contour1 = 0.68
contour2 = 0.95
do_minimal_1d_intervals = F
plot_2D_param = 0
plot_2D_num = 0
plot1 = ns omegabh2
plot2 =
num_3D_plots = 1
3D_plot1 = H0 omegam tau
triangle_plot = T
no_triangle_axis_labels = F
matlab_colscheme =
plotparams_num =0
plotparams = omegabh2 omegadmh2 tau ns
marker[nrun] = 0
limits[tau] = 0.01 N
limits[fnu] = 0 N
limits[w] = -1 N
limits[r]= 0 N
limits[asz] = 0 2
all_limits = 
finish_run_command=
force_twotail = F
PCA_num = 0
PCA_normparam = omegam
PCA_params = omegam H0 tau
PCA_func   = LLL

Antony Lewis
Posts: 1943
Joined: September 23 2004
Affiliation: University of Sussex
Contact:

Re: CosmoMC: Missing H0 in Getdist output

Post by Antony Lewis » February 08 2011

Check your input paramnames file, e.g. by default params_CMB.paramnames?

Alan Zablocki
Posts: 8
Joined: April 03 2010
Affiliation: University of Chicago, KICP

CosmoMC: Missing H0 in Getdist output

Post by Alan Zablocki » February 08 2011

Thanks for the quick reply. I have checked that file before and forgot to mention this. It does have H0, all the more why I am confused as to the missing H0. Any ideas how I can correct or test further? Thanks again.

Code: Select all

H0*           H_0		#hubble parameter is H0 km/s/Mpc
Alan

Alan Zablocki
Posts: 8
Joined: April 03 2010
Affiliation: University of Chicago, KICP

CosmoMC: Missing H0 in Getdist output

Post by Alan Zablocki » June 15 2011

This turned out to be a compiler problem! gfortran 0 : ifort 1

Post Reply