This file describes the data model for the Tilts
.
The images are written to disk as a multi-extension FITS file
prefixed by Tilts
in the Calibrations/
folder.
See :ref:`calib-naming` for the naming convention.
To examine the quality of this step the user can run the pypeit_chk_tilts script, or inspect the PNG files generated by the code. :doc:`../qa` describes how to access them.
There are currently 2 options for viewing the traced and 2D fitted tilts on the image used to construct them. Each uses the pypeit_chk_tilts script.
This is the default mode when executing, e.g.:
pypeit_chk_tilts Calibrations/Tilts_A_0_DET01.fits
The :doc:`tilt` image is shown in ginga with the traced and 2D fitted tilts over-plotted.
Here is a zoom-in screen shot for the shane_kast_red
spectrograph.
You want to check that the traced and 2D fitted tilts overlay well on the arc/OH lines.
To avoid ginga, use the --mpl flag:
pypeit_chk_tilts Calibrations/Tilts_A_0_DET01.fits --mpl
There are 3 PNG files generated per slit:
This QA shows the spectral vs. spatial positions of all
the lines traced by the code. Here is an example
from shane_kast_red
:
Here is what you hope to see in this QA:
- RMS < 0.1
- Very few red points
- The red points are mainly at the very edges of the black lines
- The black points span across the full detector
On the last point, if only a smaller portion of the detector is covered, the code is extrapolating. Indeed, the example above only shows outputs down to ~480 Spectral Pixel, which is not ideal. Of course, this will be the case for spectrographs that have data which does not span the full detector.
This QA examines the residuals of (tilt position - model)
as a function of spatial offset along the slit. Here
is an example from shane_kast_red
:
One hopes to see small residuals without any substantial spatial dependence. But, it is common for there to be significant residuals at the edges of the slit (as in this example).
This QA examines the residuals of (tilt position - model)
as a function of spetral offset along the slit. Here
is an example from shane_kast_red
:
One hopes to see:
- The black points scattering about 0
- The orange and green points not too far above 0
If one or more of your image appears to be in err, here are the things to consider:
.. todo:: Explain how to add more
Internally, the image is held in :class:`pypeit.wavetilts.WaveTilts` which subclasses from :class:`pypeit.datamodel.DataContainer`.
The datamodel written to disk is:
TILTS
content is: