hispecdrp.primitives.refpix.refpix_primitive

Classes

RefPixelCorrectionPrimitive([dataset, ...])

This step subtracts the correlated read noise (i.e., 1/f noise) using reference pixels.

class RefPixelCorrectionPrimitive(dataset=None, config=None, context=None, log_config=None, parent=None, **kwargs)[source]

Bases: BasePrimitive

This step subtracts the correlated read noise (i.e., 1/f noise) using reference pixels. Reference pixels are insensitive to light, but they track drifts in the readout electronics.

For now: This primitive applies reference pixel correction to the input data using the SCALES-DRP algorithm (Author: Athira Unni). https://github.com/scalessim/SCALES-DRP/blob/main/scalesdrp/primitives/reference.py

TODO: Eventually for HISPEC, the goal is to use 4 reference pixels on each edge of the detector as well as a 65th reference channel that is read concurrently with the 64 science channels.

TODO: explore best 1/f noise correction algorithm. We might even be able to optionally fit and subtract the 1/f noise during the spectral extraction step in stage 1.

TODO: eveluate benefits of implementing a IRS2 readout pattern for HISPEC, which has more reference pixels interleaved with science pixels.

References: Roman Roman-STScI-000673: Application and Characterization of IRRC on Roman/WFI TVAC1 Data https://www.stsci.edu/files/live/sites/www/files/home/roman/documentation/technical-documentation/_documents/Roman-STScI-000673.pdf

JWST: https://jwst-pipeline.readthedocs.io/en/stable/jwst/refpix/description.html

Parameters:
  • dataset (str or RampModel) – Input file path or data model to process. Currently, this is expected to be a raw (ramp) data product.

  • suffix (str or None, optional) – Suffix to append to the output filename. Default is 'refpix'.

  • method (str, optional) – The reference pixel correction method to use. Currently, only 'scales_drp' is supported. Default is 'scales_drp'.

  • n_channels (int or None, optional) – Number of readout channels in the detector. If None, the value is taken from the input model metadata. Default is None.

  • n_ref_left (int or None, optional) – Number of reference pixels on the left side of each channel. If None, the value is taken from the input model metadata. Default is None.

  • n_ref_right (int or None, optional) – Number of reference pixels on the right side of each channel. If None, the value is taken from the input model metadata. Default is None.

  • vertical (bool or None, optional) – Whether the readout channels are oriented vertically. If None, the value is taken from the input model metadata. Default is None.

Returns:

output (RampModel) – Reference pixel-corrected data model.

References

Unni, A., Sallum, S., Benac, P., Fitzgerald, M. P., et al., “SCALES-DRP : A Data Reduction Pipeline for an Upcoming Keck Thermal Infrared Spectrograph”, arXiv, arXiv:2509.15489 (2025) https://ui.adsabs.harvard.edu/abs/2025arXiv250915489U/abstract