Bayesian parameter estimation¶
DOWNLOAD¶
git clone -b V1.0 --depth=1 https://bitbucket.org/hanyk/bayesed BayeSED
git clone -b V1.0 --depth=1 git@bitbucket.org:hanyk/bayesed.git BayeSED
- https://bitbucket.org/hanyk/bayesed/get/V1.0.tar.bz2
USAGE¶
./bayesed (openmpi v1.8.8 is required for mpirun) [OPTIONS] -i inputfile
OPTIONS:
-a, --ann ARG1[,ARGn] Select ann model by name
e.g. -a bc03_pr_exp_ch_calzetti,iscalable
-h, -help, --help, --usage Display usage instructions
-i, --input ARG Input file containing observed photometric SEDs
e.g. -i observation/ULTRAVISTA/ULTRAVISTA0.txt
-k, --knn ARG1[,ARGn] Select knn model by name
e.g. -k bc03_pr_exp_ch_calzetti,iscalable,k,f_run
-s, --sampler ARG File containing sampling parameters for MultiNest
e.g. -s sampler.in (default)
-t, --template ARG1[,ARGn] Use template SED with the given name
e.g. -t M82,iscalable
--confidence_level ARG e.g. --confidence_level 0.5 (default)
--cosmology ARG1[,ARGn] e.g. --cosmology 70,0.7,0.3 (default)
--filters ARG e.g. --filters filter/filters.txt
--filters_selected ARG e.g. --filters_selected filter/filters_selected.txt
--IGM ARG1[,ARGn] --IGM <0|1|2>
Select the model for intergalactic medium attenuation
0 for nothing , 1 for Madau (1995) model (default), 2 for Meiksin (2006) model
--load_priors ARG load priors from .hist files for all objects
e.g.: --load_priors priors_root
--npc ARG No. of pcs to be output
e.g.: --npc 3(default)
--nzbin ARG Number of bins used to sample the prior of redshift.
e.g. --nzbin 10 (default)
--NfilterPoints ARG e.g. --NfilterPoints 30 (default)
--outdir ARG output dir for all results
e.g. --ourdir result/ (default)
--output_model_absolute_magnitude output model absolute magnitude of best fit
--output_model_apparent_magnitude output model apparent magnitude of best fit
--output_model_flux output model flux of best fit
--pca_model ARG name of .pca file
e.g.: --pca_model bc03_pr_exp_ch_calzetti
--save_bestfit Save the best fitting result for every object
--save_posteriori_sample Save the posterior sample for every object
--save_priors save priors as .hist files for all objects
--save_summary Save the file for every object
--sys_err ARG1[,ARGn] Minimum fractional error/errors for all filters
e.g. --sys_err 0 (default) or --sys_err 0.5,0.1,..,0.5
--test_priors test priors by set loglike=0
EXAMPLES¶
./bayesed -k bc03_pr_exp_ch_calzetti,1,3,1 –filters observation/ULTRAVISTA/filters.txt –filters_selected observation/ULTRAVISTA/filters_selected.txt -i observation/ULTRAVISTA/ULTRAVISTA0.txt
FAQs¶
- Q: What is the unit of photometric data in the input data file?
- A: The photometric data in the input data file must be flux and flux error in unit of uJy.
- Q: Is there any particular requirement for the order of data in the input file?
- A: The order of data in the input file must be consistent with the order of filters used in “–filters_selected observation/ULTRAVISTA/filters_selected.txt” with “iused=1”. Here, “iused=1” means there are data for the given filter in the input data file, but you can use it or not in the fitting by setting iselected=1 or 0. For example, in the example input file, we have used 30 filters, as shown by the first line “# ULTRAVISTA0 30 0”. The name of used filters is shown in the second line of the input file. So, the order of filters in observation/ULTRAVISTA/filters_selected.txt” (“Ks H J Y 8.0micron 5.8micron 4.5micron 3.6micron z+ i+ r+ Vj g+ Bj u* IA484 IA527 IA624 IA679 IA738 IA768 IA427 IA464 IA505 IA574 IA709 IA827 FUV NUV 24micron”) is the same as that in the input file(“Ks eKs H eH J eJ Y eY ch4 ech4 ch3 ech3 ch2 ech2 ch1 ech1 zp ezp ip eip rp erp V eV gp egp B eB u eu IA484 eIA484 IA527 eIA527 IA624 eIA624 IA679 eIA679 IA738 eIA738 IA767 eIA767 IB427 eIB427 IB464 eIB464 IB505 eIB505 IB574 eIB574 IB709 eIB709 IB827 eIB827 fuv efuv nuv enuv mips24 emips24”).
- Q: If some of the galaxies in the catalog don’t have data in a certain band, what should I use to indicate the absence of data to the code?
- A: You can indicate the absence of data by setting the flux and flux error to -99.
- Q: Where can I find the stellar mass of the best fit model of each galaxy?
- A: The estimation of all parameters is in a *.bayesed file. For any parameter of the model, the code gives many estimations (e.g., mass/Msol^{mean} mass/Msol^{sigma} mass/Msol^{MAL} mass/Msol^{MAP} mass/Msol^{median} mass/Msol^{lowerq} mass/Msol^{upperq}). “mass/Msol^{MAL}” is the stellar mass of the best fit model.
- Q: How to obtain photo-z with BayeSED?
- A: Set the range of allowed redshift by z_min and z_max for each galaxy in the input file. Here, different galaxy may have different z_min and/or z_max. If you have your estimation of redshift, you can set z_min=z_max=”your z_photo or z_spec”.
- Q: What is “E(B-V)” in the input file?
- A: It is about the dust extinction of the Milky Way. If it has been considered in your data, just set it to be 0.
- Q: There are some additional columns in the input file. Are they necessary?
- A: No. The input file allow some additional informations about the galaxy in columns behind the columns for the flux and flux err. These informations will be copied to the output *.bayesed file.
Copyright and Licensing¶
Copyright (C) 2014 Yunkun Han, hanyk@ynao.ac.cn
BayeSED is made freely available under the MIT license.
If you use BayeSED in your research, please cite our papers (Han, Y., & Han, Z. 2012, ApJ, 749, 123; Han, Y., & Han, Z. 2014, ApJS, 215, 2).
Changelogs¶
Version 1.0 - Aug. 2014