qso_qn-SURVEY-PROGRAM-PIXNUM.fits

Summary:

This file contains the output of QuasarNet (QSO classification algorithm and redshift fitter). When there is a disagreement between the redshift from QN and Redrock, a new redshift is fitted using only QSO templates and redshift from QN as prior.

Naming Convention:

qso_qn-SURVEY-PROGRAM-PIXNUM.fits, where SURVEY is e.g. main or sv1, PROGRAM is e.g. bright or ``dark and PIXNUM is the HEALPixel number.

Regex:

qso_qn-(cmx|main|special|sv1|sv2|sv3)-(backup|bright|dark|other)-[0-9]+\.fits

File Type:

FITS, 19 KB

Contents

Number	EXTNAME	Type	Contents
HDU0		IMAGE	Empty.
HDU1	QN_RR	BINTABLE	Output of QuasarNet afterburner.

FITS Header Units

HDU0

This HDU has no non-standard required keywords.

Empty HDU.

HDU1

EXTNAME = QN_RR

Contains the result of QuasarNet afterburner and the new redshift fit from run of Redrock with QSO templates and redshift prior from QuasarNet.

Required Header Keywords

Required Header Keywords Table

KEY	Example Value	Type	Comment
NAXIS1	151	int	width of table in bytes
NAXIS2	75	int	number of rows in table

Required Data Table Columns

Name	Type	Units	Description
TARGETID	int64		Unique target ID
RA	float64		Target Right Ascension [degrees]
DEC	float64		Target declination [degrees]
Z_NEW	float64		New redshift computed with redrock with QN prior and only qso templates
ZERR_NEW	float32		Redshift error from the new run of redrock
SV1_DESI_TARGET [1]	int64		Dark survey + calibration targeting bits for SV1
DESI_TARGET [1]	int64		Dark survey + calibration targeting bits
COEFFS	float32[10]		Coefficient of the fit for the new run of redrock
SPECTYPE	char[10]		Spectype from the redrock file
Z_RR	float32		Redshift collected from redrock file
Z_QN	float32		Redshift computed with quasarnp [2]
IS_QSO_QN_NEW_RR	logical		Is the object detected QSO with quasarnp and a new redshift fit with prior is performed?
C_LYA	float32		Confidence line for LYA (i.e.) ~ probability to be a QSO [3]
C_CIV	float32		Confidence line for CIV [3]
C_CIII	float32		Confidence line for CIII [3]
C_MgII	float32		Confidence line for MgII [3]
C_Hbeta	float32		Confidence line for Hbeta [3]
C_Halpha	float32		Confidence line for Halpha [3]
Z_LYA	float32		Redshift estimated by quasarnp with LYA line [2]
Z_CIV	float32		Redshift estimated by quasarnp with CIV line [2]
Z_CIII	float32		Redshift estimated by quasarnp with CIII line [2]
Z_MgII	float32		Redshift estimated by quasarnp with MgII line [2]
Z_Hbeta	float32		Redshift estimated by quasarnp with Hbeta line [2]
Z_Halpha	float32		Redshift estimated by quasarnp with Halpha line [2]

[1] (1,2)

Optional

[2] (1,2,3,4,5,6,7)

Z_QN is the redshift estimated on the line of the highest confidence

[3] (1,2,3,4,5,6)

The QN selection is performed with these parameters. As it stands, in QN afterburner everything with np.max(confindence) > 0.5 is considered as a quasar. However, specific cut will be used depends on each target class; QSO_target will use np.max(confidence) > 0.95. See: https://github.com/echaussidon/desispec/blob/720153babcf85dd93530252b0c1f631d48edfc0d/bin/desi_qso_qn_afterburner#L236

Notes and Examples

These files are generated with https://github.com/desihub/desispec/blob/master/bin/desi_qso_qn_afterburner