API¶
This page details the methods and classes provided by the kmos_tools module.
Exposure Class¶
Class to hold individual KMOS exposures for post-processing
Takes *SCI_RECONSTRUCTED.fits filenames as input
- Usage:
Create a instance of the exposure class for each exposure file:
from kmos_tools import exposure fname = 'KMOS_SCI_RECONSTRUCTED.fits' exp = exposure.Exposure(fname)
-
class
kmos_tools.exposure.Exposure(reconstructed_fits_path, vb=False)[source]¶ A single KMOS exposure object
Takes SCI_RECONSTRUCTED filename as input, to be used for other operations.
-
hdulist¶ HDUlist of IFUs in the exposure
- Type
astropy sequence of HDU objects or single HDU
-
hdr¶ FITS header
- Type
astropy header
-
Sky residual cleaning¶
Module to clean up sky subtraction residuals in KMOS data
Works on individual exposures which must then be combined via kmos_combine.
-
kmos_tools.sky_clean.make_sky_residual_spectra(exposure, use_named_targets=[''], clobber=True, plot=True)[source]¶ Calculate all the sky residual corrections for one DIT ~ a la Trevor Mendel
Generate median 1D sky spectra for each detector, only from certain cubes, and save them to use for subtraction. E.g. for KLASS we set use_named_targets=[‘S1’, ‘S3’] to only use S1 and S3 targets (high z ~ empty).
- Parameters
exposure (object) – exposure object
use_named_targets (list) – list of string to match with target name, uses only IFUs containing those targets to make sky residual spectra, default uses all IFUs
clobber (bool) – Make a new sky spectrum if it already exists
plot (bool) – make and save a plot of the median sky spectra
- Returns
Saves FITS file with 1D sky spectra for each detector
-
kmos_tools.sky_clean.subtract_sky_residual_spectra(exposure, clobber=True, plot=True)[source]¶ Subtract residual sky spectrum from each IFU using relevant detector sky spectrum
Rescales 1D sky spectrum \(s_\lambda\) in each spatial pixel by \(A_{x, y}\) for each IFU such that sky subtracted spectrum:
\[f_{x, y, \lambda}^\mathrm{skysub} = f_{x, y, \lambda} - A_{x, y}s_\lambda\]Where \(A_{x, y}\) is obtained by minimising:
\[\left( \frac{f_{x, y, \lambda}^\mathrm{skysub}}{\sigma_\lambda} \right)^2\]
-
kmos_tools.sky_clean.sky_residual(params, sky, data, err)[source]¶ Residual for 1D sky-scaling
Used to find optimal rescaling of 1D spectrum, A, for each spatial pixel
\[\begin{split}R_i = \frac{(f_i - A s_i)}{\sigma_i} \\ \chi^2 = \sum_i R_i^2\end{split}\]- Parameters
params (params) – parameters
sky (ndarray) – 1D sky spectrum
data (ndarray) – 3D data cube
err (ndarray) – 1D error spectrum
- Returns
residual
Finding star offsets¶
Module to find stars in an exposure, and fit their spatial positions to use in kmos_combine --user_shifts
Usage:
# List of frames to check (should be all the frames you want to combine)
frame_list = ['frame1.fits', 'frame2.fits']
fname_stars = 'stars.txt' # File to save star parameters to
fname_combine = 'combine.sof' # .sof file to create
kt.star_positions_batch(frame_list, psf_cut=0.8, starparams_filename=fname_stars, combinefiles_filename=fname_combine)
fname_usershifts = 'usershifts.txt' # file to save usershifts to
kt.make_user_shifts_file(starparams_filename=fname_stars, usershifts_filename=fname_usershifts)
-
kmos_tools.star_offsets.make_user_shifts_file(starparams_filename, usershifts_filename=None)[source]¶ Create file of user shifts to use with
kmos_combine --method="user"Note
To run
kmos_combineyou must use the frames in the order they are used here i.e. so the offset matches the right frame
-
kmos_tools.star_offsets.star_positions_batch(frame_list, psf_cut=0.8, edge_x=2.0, edge_y=2.0, star_ifu=None, starparams_filename=None, combinefiles_filename=None)[source]¶ Given a list of exposure file names, will find stars and measure PSFs.
Given a list of exposure file names, will find stars and measure PSFs. Then creates a list of frames with PSF FWHM below a given value, saves the list of star positions (to calculate user shifts) and creates a .sof file of the ‘good’ frames.
- Parameters
frame_list (list) – List of file names of individual
SCI_RECONSTRUCTED.fitsfilespsf_cut (float) – max PSF to use for combining (in arcsec) [default = 0.8 arcsec]
edge_x (float) – x_star > edge_x to include (i.e. not on the edge) [default = 2 pixels]
edge_y (float) – y_star > edge_y to include (i.e. not on the edge) [default = 2 pixels]
star_if (int) – IFU star is on
- Yields
starparams_filename (str) – table with parameters of good stars combinefiles_filename (str): .sof file with list of good frames
-
kmos_tools.star_offsets.find_star_ifu(exposure)[source]¶ Find the IFUs in the SCI_RECONSTRUCTED cubes containing a star
-
kmos_tools.star_offsets.star_fit_profile(exposure)[source]¶ Fit gaussian profiles to star to find PSFs
Using ESO command line pipeline tools (
esorex) extract the star from the exposure, fit a Gaussian profile to it, and save the fitted star to a new fits filestar_file_name.
-
kmos_tools.star_offsets.star_psf(exposure, clobber=True, vb=False)[source]¶ Get the PSF profiles of the star for calculating user shifts
Read the file
star_file_nameand return the main parameters- Parameters
exposure (object) – exposure object
- Returns
PSF centroid x in pixels psf_center_y (float): PSF centroid y in pixels psf_fwhm (float): PSF FWHM in arcsec psf_ba (float): PSF BA # TODO I don’t remember what this is!! psf_pa (float): PSF position angle in degrees exposure.invert (bool): is the flux weird?
- Return type
psf_center_x (float)
Pipeline fixes¶
Quick fixes for KMOS pipeline problems. Methods are:
-
kmos_tools.pipeline_fixes.rotation_fix()[source]¶ Rotates frames so that they are all orientated in the same direction for combining
-
kmos_tools.pipeline_fixes.calib_fix()¶ Fix flux calibration if it wasn’t done by the pipeline (i.e. no standard observed on on detector)
-
kmos_tools.pipeline_fixes.flux_fix()[source]¶ Fixes an old (now fixed??) pipeline problem where flux needs to be divided by DIT length in seconds
-
kmos_tools.pipeline_fixes.rotation_fix(exposure, updateheader=False, keepsize=False)[source] Sometimes the instrument was rotated, this rotates all the frames to the same angle for combining
For some OBs the field is rotated so North is no longer up. This is when ocs.rot.offangle != 0
Need to rotate back to North = up so we can combine nicely
Check if frame is rotated
Use kmo_rotate to rotate the frame
Update the headers
Rename old file to _rotoffangle###
Save file
Note
Best to run this on sky sub files (_noS2.fits)
-
kmos_tools.pipeline_fixes.flux_fix(exposure, clobber=True)[source] Fix Reflex exposure time bug
Fix the bug in Reflex by dividing the calibrated flux from the standard stars by the DIT for science exposures
Note
This is now obsolete as fixed in kmos pipeline v. > 1.4