Data Products
Generated output files
Depending on user configuration, nirwals_reduce generates multiple output files. These are typically named after the filename of the input file, followed by a standardized prefix to identify the data product:
*_reduced.fits
These are the reduced files, and are the only files generated in any case. See below for details on FITS extensions and data units
*__stack_raw.fits
are full-cube output files containing all input reads combined into a single data cube. The data in this file is straight from the detector, with no calibrations applied. Pixels brighter than the configured saturation level are masked as NaNs.
*__stack_refpixcorr.fits
Similiar to the _stack_raw.fits, but after applying the per-read reference pixel correction.
*__stack_linearized.fits
As above, but with reference pixel AND non-linearity correction applied. This is the final stage of data preparation used for the up-the-ramp fitting.
Output file extensions
The final reduced files contain a number of output extensions that contain results from different methods or metadata describing and/or derived from and as part of the data processing.
PRIMARY
The primary FITS header does not contain any image data, but rather all headers relevant to describe the input and output data. Most of the FITS headers are taken from the selected reference exposure (typically the first in the read sequence), but also contains a number of keywords summarizing the entire sequence. See below for more details.
SCI:
This is the final calibrated slope derived from the up-the-ramp fitting.
UNITS Data is given in data numbers (counts) per second, with NO gain correction.
MEDIAN:
tbd
NOISE
tbd
NPAIRS
MAX_T_EXP
Per-pixel maximum integration before exceeding the saturation level.
PROVENANCE
Unlike all other extensions, this extension is a FITS table, with key/value pairs describing many aspects relevant for reproducing the output file. This includes:
List of input files
Calibration products used (e.g. non-linearity corrections)
System configuration
Package version information for the nirwals package itself and other important packages (such as numpy & astropy)