betatree

Decision tree pipeline for p(z) estimation from CFHT/Megacam photometry [arXiv:1610.01160]

Installation

Clone this repository with git clone git@github.com:danielgruen/betatree.git then cd betatree

You likely have to edit the Makefile to match the path to your installation of the CCfits library.

Then call make to build the betatree executable and download the required reference catalogs.

Usage

Call ./betatree [zcluster] [output tree file] [maximum i band depth] [bands_keys band_depths] to build a color-magnitude tree and write it to file, e.g. ./betatree 0.5 tree.fits 25 g 25.5 r 25.5 z 26 to generate a tree optimized for lenses at z=0.5, for a magnitude limited sample of galaxies with 20<i<25, using color information from i band data and g, r, z band data down to magnitudes 25.5, 25.5, 26. We recommend choosing the depths such that your photometric signal-to-noise ratio is at least 10 for galaxy brighter than these limits. The resulting tree can be used for galaxies with 20<i<25 and any g, r, z flux.

Photometric information from the reference catalogs is complete and valid down to the magnitude limits described in arXiv:1610.01160. All magnitudes are MAG_AUTO (approximately equal to the total magnitude) in the AB system for CFHT/Megacam filters, with the old i band. The new i band is available and called 'y'.

Output

The FITS file output contains three extensions.

Tree information

• BETATREE_BRIGHTCUT,BETATREE_MAXDEPTH: lower and upper limit in detection band (=i) magnitude for which tree is valid
• BETATREE_DEPTH_*: depth in other bands up to which splits are made in the tree; should be at a SNR>=10 in your data
• BETATREE_Z_LENS: lens redshift for which the tree was optimized
• BETATREE_BETAMEAN_DEEP: lensing-weighted mean \beta=D_ds/D_s for the CFHTLS-DEEP magnitude-limited sample, estimated from the tree with 8-band template-fit photo-z
• BETATREE_BETAMEAN_D2,BETATREE_BETAMEAN_C2015: lensing-weighted mean \beta from D2 (COSMOS) only and from the Laigle+2015 COSMOS photo-z, indicative of overall biases; 0 if i band limit is above 24.7, where C2015 is not quite complete
• BETATREE_RELATIVE_COSMIC_STDEV: relative standard deviation \sigma_\beta/\beta of the lensing-weighted mean \beta estimated from variance of the four DEEP fields
• BETATREE_POFZ_*: description of the grid of the p(z) (in equal steps in z) saved in the third extension
• BETATREE_NLEAVES: number of leaves in tree
• BETATREE_NMIN: minimum number of reference galaxies per leaf
• BETATREE_VERSION: version information of the betatree code used

Leaf information

For each leaf, a line with

• leafid: running index
• nobj: number of objects in the CFHTLS-DEEP magnitude limited sample in leaf
• i_min/max: i band range of leaf
• ?i_min/max: range of ?-i color of leaf
• beta,sigmabeta: mean value and uncertainty of mean of \beta of all CFHTLS galaxies in leaf
• DdsDs????_mean: mean value of \beta without the noise bias correction described in arXiv:1610.01160
• DdsDssq????_mean: mean value of \beta^2
• background???_mean,cluster????_mean: probability of a galaxy in this leaf to be in the background z>1.06*(1+z_l) / in the cluster redshift slice |z-z_l|<0.06*(1+z_l)
• beta_noD[1234]: mean \beta estimated without DEEP field 1/2/3/4, useful for generating cosmic variance metric with jackknife

p(z)

For each leaf, a line with

• leafid: running index as in previous table
• pofz: array of normalized p(z) with elements i=1...400, where element i describes z=i/100\pm0.005

Help

I would be happy to help you use this code and data correctly -- please contact me at dgruen@stanford.edu.

Acknowledgements

The method is described in detail in Gruen & Brimioulle (2017) and first applied in Cibirka et al. (2017). Publications making use of this code should also acknowledge the public CFHTLS-DEEP and WIRDS data and the Photoz code (Bender et al. 2001) and COSMOS photo-z (Laigle et al. 2015) used in the reference catalogs. Support for DG was provided by NASA through the Einstein Fellowship Program, grant PF5-160138.