Data model: BOSS Target Files

Overview

General description: Boss target files contain targeting info for the galaxies, quasars, and standards. The basic bosstarget files match one-to-one with the data sweep (calibObj) from which they were generated. Thus there is one target file per RUN/CAMCOL for the given set of RERUNs. These files at a minimum contain the run/rerun/camcol/field/id of each object and the boss_target flag bits, and a separate HDU containing status information about the processing. There are also collated files, which are reduced from the target files based on target flags and other information and collated with the full information from the calibObj from which they were generated.

Naming convention:

  1. $BOSS_TARGET/$TARGET_RUN/bosstarget-$TARGET_TYPE-$RUN-$CAMCOL-$RERUN-$TARGET_RUN.fits
  2. $BOSS_TARGET/$TARGET_RUN/bosstarget-$TARGET_TYPE-$TARGET_RUN-collate.fits
  3. $BOSS_TARGET/$TARGET_RUN/bosstarget-$TARGET_TYPE-$TARGET_RUN-collate-$CHUNK.fits

for TARGET_TYPE in "lrg" "qso" "std". TARGET_RUN is generally a date string with some modifications, e.g. 2009-07-15c where c means "commissioning". The format for $RUN, $RERUN, $CAMCOL is the usual I6, string, and I1 respectively.

As shown above, the directory may contain intermediate "chunk" collated files written out during the parallel processing. The $CHUNK variable has format I3. These files are not critical and may disappear.

Relevant IDL classes and methods:

  1. boss/bosstarget/pro/bosstarget__define.pro
  2. boss/bosstarget/pro/bosstarget_$TARGET_TYPE__define.pro

    for TARGET_TYPE in "lrg" "qso" "std"

Basic Target File Data Model

Header Keywords

There are no required header keywords. The details of the processing are instead contained in the first binary table hdu.

First HDU: The Status Structure

This is the first binary table hdu (1 for zero-offset extension counting). This hdu contains details and status about the processing and is generally referred to as the "status structure".

Required Fields

Second HDU: Target Selection Information

In principle only the run, rerun, camcol, field, id, and boss_targetN are absolutely required. This is because these files are designed to be collated with the original calibObj files when generating a target list.

Required Fields

In practice there can be many more fields in the output file, especially for the QSOs. See the details listed below.

Specifics for Different Target Types

Quasars

The qso selection is performed on the "star" calibObj files.

Galaxy

The galaxy selection is performed on the "gal" calibObj files.

Standards

The standards selection is performed on the "star" calibObj files.

Collated Target File Data Model

Header Keywords

There are no required header keywords. The details of the processing are instead contained in the first binary table hdu.

First HDU: The Collated Status Structure

This is the first binary table hdu (1 for zero-offset extension counting). This hdu contains details and status about the processing and is generally referred to as the "status structure".

Required Fields

Second HDU: Target Selection Information and Full calibObj Data

In principle only the run, rerun, camcol, field, id, boss_target selection flags, RA, DEC, and fiber magnitudes are absolutely required since these are required by the tiling. In practice the union of the outputs from the basic target files with the entire set of information from the calibObj files is present in the collated files. As explained above, the different target types use different calibObj types ('star' or 'gal') as their basis. Plese see the documentation for the calibObj files for more info on what fields are available.

Required Fields