redvsblue

"redshift versus blueshift"
Quasar and emission line precise redshift fitting from prior

Presentation

This package aims at measuring a precise redshift given a broad redshift prior. To do so the package, for each emission line or the full spectrum:

Runs a coarse chi2 scan as a function of redshift, using the input PCA+boradband Legendre polynomials
Finds three local minima
Does a finer chi2 scan in each minima
Defines the global PCA redshift, "ZPCA" from the best minimum of the three. ZPCA is a redshift estimator biased toward the computation of the PCA.
The redshift of the line, "ZLINE", is defined from the maximum of the best-fit model of the line. ZLINE is a redshift estimator un-biased toward the velocity of the line, but can be biased with respect to the cosmological redshift.

Chi2 scan vs. redshift	Best fit

Install

To install simply clone the repository, add to your PATH and to your PYTHONPATH

git clone https://github.com/londumas/redvsblue
PYTHONPATH=$PYTHONPATH:<path_to_folder>/redvsblue/py/
PATH=$PATH:<path_to_folder>/redvsblue/bin/

Run

To get all available options:

redvsblue_lineFitter.py --help

To run on SDSS data, do:

redvsblue_lineFitter.py
--out <path_to_write_output>.fits
--in-dir $BOSS_SPECTRO_REDUX/v5_13_0/
--drq DR12Q_v2_10.fits
--z-key Z_VI
--qso-pca <path_to_folder>/redvsblue/etc/rrtemplate-qso.fits

To run on DESI data, for example:

redvsblue_lineFitter.py
--out <path_to_write_output>.fits
--in-dir /project/projectdirs/desi/datachallenge/redwood/spectro/redux/redwood/spectra-64/
--drq /project/projectdirs/desi/datachallenge/redwood/spectro/redux/redwood/zcatalog-redwood-target-truth.fits
--z-key Z
--qso-pca <path_to_folder>/redvsblue/etc/rrtemplate-qso.fits
--data-format DESI
--no-extinction-correction

Output

The output is a FITS file, with one HDU per redshift type:

CAT, HDU=1: redshift prior
PCA, HDU=2: PCA redshift
HALPHA, HDU=3: redshift of the HALPHA line
HBETA, HDU=4: redshift of the HBETA line
MGII, HDU=5: redshift of the MGII line
CIII, HDU=6: redshift of the CIII line
CIV, HDU=7: redshift of the CIV line
LYA, HDU=8: redshift of the LYA line

For each best fit readshift, the code gives the following quantities:

ZLINE: best fit redshift of the line according to maximum of PCA
ZPCA: best fit redshift according to PCA
ZERR: redshift error
ZWARN: redshift warning, use ZWARN=0 for reliable redshifts
CHI2: chi^2 of the best fit
DCHI2: Delta chi^2 against a Legendre polynomial, i.e. gives the significance of the redshift
NPIXBLUE: number of pixel on the blue side around the prior redshift
NPIXRED: number of pixel on the red side around the prior redshift
NPIX: total number of pixels around the prior redshift
NPIXBLUEBEST: number of pixel on the blue side around the best redshift
NPIXREDBEST: number of pixel on the red side around the best redshift
NPIXBEST: total number of pixels around the best redshift
SNR: mean signal-to-noise ratio per pixel in the spectral region used for measuring the redshift

Why Red vs. Blue

RED vs BLUE, from Corridor: https://www.youtube.com/watch?v=arg_aHzviQw

Release follow the name of the queen of France

https://fr.wikipedia.org/wiki/Liste_des_reines_et_imp%C3%A9ratrices_de_France

londumas / redvsblue