"""pixtopix - A module to perform coordinate transformation from pixel
coordinates in one image to pixel coordinates in another frame
:Authors: Warren Hack
:License: :doc:`/LICENSE`
"""
import numpy as np
from stsci.tools import fileutil
from stwcs import wcsutil, distortion
from . import wcs_functions
from . import util
from . import __version__
__all__ = ['tran']
__taskname__ = 'pixtopix'
[docs]
def tran(inimage, outimage, direction="forward", x=None, y=None, coordfile=None,
colnames=None, separator=None, precision=6, output=None, verbose=True):
"""Primary interface to perform coordinate transformations in pixel
coordinates between 2 images using STWCS and full distortion models
read from each image's header.
Parameters
----------
inimage : str
Full filename with path of input image, an extension name ``['sci',1]``
should be provided if input is a multi-extension FITS file.
outimage : str, optional
Full filename with path of output image, an extension name ``['sci',1]``
should be provided if output is a multi-extension FITS file.
If no image gets specified, the input image will be used to generate a
default output WCS using ``stwcs.distortion.util.output_wcs()``.
direction : str
Direction of transform (forward or backward). The 'forward' transform
takes the pixel positions (assumed to be from the 'input' image) and
determines their position in the 'output' image. The 'backward'
transform converts the pixel positions (assumed to be from the
'output' image) into pixel positions in the 'input' image.
x : float, optional
X position from image
y : float, optional
Y position from image
coordfile : str, optional
Full filename with path of file with starting x,y coordinates
colnames : str, optional
Comma separated list of column names from 'coordfile' files
containing x,y coordinates, respectively. Will default to
first two columns if None are specified. Column names for ASCII
files will use 'c1','c2',... convention.
separator : str, optional
Non-blank separator used as the column delimiter in the coordfile file
precision : int, optional
Number of floating-point digits in output values
output : str, optional
Name of output file with results, if desired
verbose : bool
Print out full list of transformation results (default: False)
Returns
-------
outx : float
X position of transformed pixel. If more than 1 input value, then it
will be a numpy array.
outy : float
Y position of transformed pixel. If more than 1 input value, then it
will be a numpy array.
Notes
-----
This module performs a full distortion-corrected coordinate transformation
based on all WCS keywords and any recognized distortion keywords from the
input image header.
It can be called from within Python using the syntax::
>>> from drizzlepac import pixtopix
>>> outx,outy = pixtopix.tran("input_flt.fits[sci,1]",
... "output_drz.fits[sci,1]", "forward", 100, 100)
Examples
--------
1. The following command will transform the position 256,256 from
'input_flt.fits[sci,1]' into a position on the output image
'output_drz.fits[sci,1]' using::
>>> from drizzlepac import pixtopix
>>> outx,outy = pixtopix.tran("input_file_flt.fits[sci,1]",
... "output_drz.fits[sci,1]", "forward", 256, 256)
2. The set of X,Y positions from 'output_drz.fits[sci,1]' stored as
the 3rd and 4th columns from the ASCII file 'xy_sci1.dat'
will be transformed into pixel positions from 'input_flt.fits[sci,1]'
and written out to 'xy_flt1.dat' using::
>>> from drizzlepac import pixtopix
>>> x,y = pixtopix.tran("input_flt.fits[sci,1]", "output_drz.fits[sci,1]",
... "backward", coordfile='xy_sci1.dat', colnames=['c3','c4'],
... output="xy_flt1.dat")
"""
single_coord = False
if coordfile is None:
if isinstance(x, np.ndarray):
xlist = x.tolist()
ylist = y.tolist()
elif not isinstance(x, list):
xlist = [x]
ylist = [y]
single_coord = True
else:
xlist = x
ylist = y
else:
if colnames in util.blank_list:
colnames = ['c1', 'c2']
# Determine columns which contain pixel positions
cols = util.parse_colnames(colnames, coordfile)
# read in columns from input coordinates file
xyvals = np.loadtxt(coordfile, usecols=cols, delimiter=separator)
if xyvals.ndim == 1: # only 1 entry in coordfile
xlist = [xyvals[0].copy()]
ylist = [xyvals[1].copy()]
else:
xlist = xyvals[:, 0].copy()
ylist = xyvals[:, 1].copy()
# start by reading in WCS+distortion info for each image
im1wcs = wcsutil.HSTWCS(inimage)
if im1wcs.wcs.is_unity():
print("####\nNo valid input WCS found in '{}'."
"\n Results may be invalid.\n####\n".format(inimage))
if util.is_blank(outimage):
fname, fextn = fileutil.parseFilename(inimage)
numsci = fileutil.countExtn(fname)
chips = [wcsutil.HSTWCS(fname, ext=('sci', e + 1)) for e in range(numsci)]
if len(chips) == 0:
chips = [im1wcs]
im2wcs = distortion.utils.output_wcs(chips)
else:
im2wcs = wcsutil.HSTWCS(outimage)
if im2wcs.wcs.is_unity():
print("####\nNo valid output WCS found in '{}'."
"\n Results may be invalid.\n####\n".format(outimage))
# Setup the transformation
p2p = wcs_functions.WCSMap(im1wcs, im2wcs)
if direction[0].lower() == 'f':
outx, outy = p2p.forward(xlist, ylist)
else:
outx, outy = p2p.backward(xlist, ylist)
if isinstance(outx, np.ndarray):
outx = outx.tolist()
outy = outy.tolist()
# add formatting based on precision here...
xstr = []
ystr = []
for ox, oy in zip(outx, outy):
xstr.append('{0:.{1}f}'.format(ox, precision))
ystr.append('{0:.{1}f}'.format(oy, precision))
if verbose:
print('# Coordinate transformations for ', inimage)
print('# X(in) Y(in) X(out) Y(out)\n')
for xs, ys, a, b in zip(xlist, ylist, xstr, ystr):
print("%.4f %.4f %s %s" % (xs, ys, a, b))
# Create output file, if specified
if output:
with open(output, mode='w') as f:
f.write("# Coordinates converted from %s\n" % inimage)
for xs, ys in zip(xstr, ystr):
f.write('%s %s\n' % (xs, ys))
print('Wrote out results to: ', output)
if single_coord:
outx = outx[0]
outy = outy[0]
return outx, outy
# --------------------------------
# TEAL Interface functions
# (these functions are deprecated)
# ---------------------------------
def run(configObj):
coordfile = util.check_blank(configObj['coordfile'])
colnames = util.check_blank(configObj['colnames'])
sep = util.check_blank(configObj['separator'])
outfile = util.check_blank(configObj['output'])
outimage = util.check_blank(configObj['outimage'])
verbose = configObj['verbose'] or outfile is None
tran(configObj['inimage'], outimage, direction=configObj['direction'],
x=configObj['x'], y=configObj['y'], coordfile=coordfile,
colnames=colnames, separator=sep, precision=configObj['precision'],
output=outfile, verbose=configObj['verbose'])