Large-scale Structure Samples

The Large-Scale Structure samples built out of a VAGC version are stored under the directory tree: 

$LSS_REDUX/dr6fix

At the bottom of this page we describe the standard ASCII format files that have been in place for SDSS users since sample7 or so.

Global files

In the LSS directory are the files:

  1. lss_combmask.dr6fix.fits: the geometry of the survey
  2. lss_index.dr6fix.fits: positions of all objects in VAGC and which lss_geometry polygon they are in (-1 if outside OR if inside the bright star mask)
There are some auxiliary files which are occasionally useful:
  1. lss_geometry.dr6fix.fits: file describing geometry of the full large-scale structure sample
  2. lss_bsmask.dr6fix.fits: file describing the bright star mask
  3. lss_imta.dr6fix.fits: expresses which "best" fields intersect which "target" fields
There is also an ASCII directory with:
  1. lss_geometry.dr6fix.ply: mangle v1.4 format file describing the geometry
  2. lss_bsmask.dr6fix.ply: mangle v1.4 format file describing the bright star mask
  3. lss_geometry_info.dr6fix.dat: information for each polygon in the .ply file
  4. lss_bsmask_info.dr6fix.dat: information for each polygon in the .ply file

The ASCII directory also has some files that have some of the imaging information. For example:

  1. shapes_r.dr6fix.dat: shape information

Random catalogs

Random catalogs distributed with an equal surface density within the window (and outside the mask) can be found in:

$LSS_REDUX/dr6fix/random/lss_random-#.dr6fix.fits

There are ASCII versions of these files as well. Finally, copies of these random files made for various subsamples (where we have calculated completenesses and flux limits) are described below.

Inverse random catalogs are occasionally useful, so that you can toss down a sphere and ask how much of it is outside the mask. One can find these at

$LSS_REDUX/dr6fix/random/lss_invran-#.dr6fix.fits

There are ASCII versions of these files as well. Finally, copies of these random files are made for various subsamples (where we have calculated completenesses and flux limits).

Data subsamples -- "pre-redshift" selection

In addition there are a number of subdirectories which are appropriate for large-scale structure in the Main sample. Each subdirectory corresponds to a "pre-redshift" selection criterion; this means that the objects and the area of sky have been selected according to flux limit, completeness, and other properties which do not require knowing the object redshift. The correspondence between the subdirectory name ("letter") and these criteria are given in a Yanny parameter file found here. These subdirectories are of the form: 

$LSS_REDUX/dr6fix/[letter]

The pre-redshift selection criteria are:

  1. mmin: bright limit
  2. mmax: faintest faint limit (that is, don't select fainter than this --- the actual faint limit in any direction might be brighter than this value)
  3. minsampling: minimum sampling of a sector to include
  4. psfmodel: star-galaxy separation criteria (smaller includes more stellar-like objects)
  5. r50bright, magbright: objects brighter than magbright with Petrosian R50 values smaller than r50bright (in arcsec) are excluded (these tend to be close binary stars)
  6. collision_type: algorithm used to fix collisions (see description)
  7. flux_type: type of flux used for flux limit (usually "petro")
  8. band: band used for flux limit (indexing ugriz, usually "2" for "r")
  9. minscore: minimum photometric score for the target version of a field for this to be included (the "best" version will always be equal or better). Basically, this should be >0.5, but for some purposes (like finding voids) we don't care about photometricity as much as contiguity.

If you are doing large-scale structure statistics and want to include "collision corrections" then you want to use one of:

  1. full: everything, following the variable flux limit to get the most number of galaxies
  2. safe: everything, using a constant flux limit of 17.6 for simplicity, and a bright limit of 14.5
  3. bright: like safe, without the bright limit
  4. void: like safe, but including areas with questionable photometry for the sake of contiguity

If you are doing galaxy property statistics, you probably don't want the collision corrections, so you want one of:

  1. all: like full, but no collision corrections
  2. bsafe: like safe, but no collision corrections
  3. bbright: like bright, but no collision corrections
  4. bvoid: like void, but no collision corrections

The files inside each directory are:

  1. letter_catalog.[sample][letter].fits: has which set of objects passes the "pre-redshift" cuts
  2. combmask.[sample][letter].fits: describes window of this sample (this set of polygons is smaller than that in lss_combmask because we unify polygons which have the same flux limit and completeness).
  3. window.[sample][letter].fits: describes window of this sample without the mask
  4. mask.[sample][letter].fits: mask to exclude for this sample
  5. lss_to_window.[sample][letter].fits: for each polygon in lss_geometry, this points to which polygon in window.fits contains it.
  6. random/random-#.[sample][letter].fits: files with random catalogs distributed with constant surface density on the sky

Data subsamples -- "post-redshift" selection

Further subdirectories provide complete subsamples based on cuts made after the redshift determination: on redshift, luminosity, intrinsic color, etc. The correspondence between the subdirectory name ("post") and these criteria are given in a Yanny parameter file found here. These subdirectories are of the form: 

$LSS_REDUX/dr6fix/[letter]/[post]

The post-redshift selection criteria are:

The redshift evolution model is:

M(z) = M(z=qz0) - q0 [1 + q1 (z - qz0)] (z-qz0)
The nonlinearity is essential to getting the number counts correct at high redshift.

The files inside each post-redshift directory are:

  1. post_catalog.[sample][letter][post].fits: has which set of objects passes the "post-redshift" cuts

There is a subdirectory "vmax" which has information about the maximum volume over which you can over a galaxy. It contains files like:

  1. vmax-q2.80a-1.80.dr6fixall1.fits: Vmax estimates
  2. vmax-noevol.dr6fixall1.fits: Vmax estimates (for no-evolution case)

You may see some other subdirectories. Basically, you should ignore them if they are not documented here, because I might be experimenting --- you don't want to use something that is changing/wrong.

LSS ASCII subsamples

We provide some ASCII subsamples where we have dealt with the radial selection function, etc.

Many people just need ASCII files describing the sample and the geometry. We provide these in directories of the form: 

$LSS_REDUX/[sample]/[letter]/[post]/lss

They are described almost completely by the README files that they contain.

They additionally have SDSSPix descriptions of their geometry provided by Ryan Scranton.

The formats have changed slightly since sample12. Mainly:

  1. they contain mangle v1.4 ASCII files describing the window and mask
  2. as implied by the above, they have bright star mask information as well as window information
  3. ASCII random catalogs are now stored in the directories: 
    $LSS_REDUX/[sample]/[letter]/random
    (because they are all the same for all "post-redshift" subsamples).
  4. they no longer refer to "boundaries" but instead to "mregions", which have constant flux limits but are not (necessarily) contiguous on the sky

NYU Value-Added Galaxy Catalog

Please contact us with comments or questions.