coadd_centroid

sofia_redux.instruments.forcast.register.coadd_centroid(data, reference, header=None, variance=None, crpix=None, border=0, rot_angle=None, missing=nan, shift_order=None, rotation_order=1, get_offsets=False)[source]

Shift an image for coadding using a centroid algorithm

Steps are:

  1. Rotate data and variance wrt reference (if rot_angle is provided)

  2. Run peakfind on data wrt reference on 1 peak

  3. Return rotated data (if it was rotated) should peakfind fail.

  4. Shift image by offsets calculated wrt reference using peakfind.

  5. Record the shifts in the header with the COADX0, COADY0 keywords.

Parameters:
datanumpy.ndarray

input image array (nrow, ncol)

referencenumpy.ndarray

input reference array to compare data to (nrow, ncol)

headerastropy.fits.header.Header

FITS header of input data

borderint

exclude border pixels from the outside edges when attempting to find peaks in the image and reference

shift_orderint, optional

Order of interpolation for the shift. The shift order must be between 0 and 5, with a default of 3

rotation_order: int, optional

Order of interpolaton for the rotation. The rotation order must be between 0 and 5, with a default of 3

variancenumpy.ndarray

propagate provided variance if set (nrow, ncol)

rot_anglefloat

Indicates that the data image is rotated wrt the reference image by this amount. The data image (and variance if supplied) will be rotated clockwise by rot_angle degrees

crpixlist, optional

If provided, will be updated to the new [CRPIX1, CRPIX2] values following the shift

get_offsetsbool, optional

Do not shift the image. Only return the (x, y) offset

missingfloat, optional

Value to represent missing data during shift

Returns:
2-tuple

numpy.ndarray : The shifted image (nrow, ncol) numpy.ndarray : The shifted variance (nrow, ncol)