Chess-like pattern when using rountrip_coords with GWCS, when using dask

[astrofrog]

I am seeing a very strange issue. The following example tries to reproject a dataset with a GWCS to a slightly different GWCS:

import os
import asdf
from copy import deepcopy
from astropy import units as u
from reproject import reproject_interp
import matplotlib.pyplot as plt

aia = asdf.open(
    os.path.join(
        "/Users/tom/Code/Astropy/reproject/reproject/tests/data/aia_171_level1.asdf"
    )
)
input_data = (aia["data"][...], aia["wcs"])
wcs_out = deepcopy(aia["wcs"])
wcs_out.forward_transform.offset_0 = -60.3123 * u.pix
wcs_out.forward_transform.offset_1 = -61.9422 * u.pix
shape_out = aia["data"].shape

array1 = reproject_interp(
    input_data,
    wcs_out,
    shape_out=shape_out,
    return_footprint=False,
)

array2 = reproject_interp(
    input_data,
    wcs_out,
    shape_out=shape_out,
    return_footprint=False,
    return_type="dask",
    block_size=(200, 200),
)
array2 = array2.compute(scheduler="synchronous")

fig = plt.figure()
ax1 = fig.add_subplot(1, 2, 1)
ax1.imshow(array1, origin="lower")
ax2 = fig.add_subplot(1, 2, 2)
ax2.imshow(array2, origin="lower")
fig.savefig('chess.png')

The result is:

This goes away if using rountrip_coords=False so is related to the round-tripping. However this seems to happen only when using dask, even if just in synchronous mode. Furthermore the chess-like pattern is unrelated to the chunk size. Even if I set the chunk size to be larger than the image, so that there is only one chunk, the chess-like pattern is present.

[astrofrog]

Ok so this is not related to dask per se but to the use, I think, of SlicedLowLevelWCS:

import os
import asdf
from copy import deepcopy
from astropy import units as u
from reproject import reproject_interp
import matplotlib.pyplot as plt
from astropy.wcs.wcsapi import BaseHighLevelWCS, SlicedLowLevelWCS
from astropy.wcs.wcsapi.high_level_wcs_wrapper import HighLevelWCSWrapper

aia = asdf.open(
    os.path.join(
        "/Users/tom/Code/Astropy/reproject/reproject/tests/data/aia_171_level1.asdf"
    )
)
input_data = (aia["data"][...], aia["wcs"])
wcs_out = deepcopy(aia["wcs"])
wcs_out.forward_transform.offset_0 = -60.3123 * u.pix
wcs_out.forward_transform.offset_1 = -61.9422 * u.pix
shape_out = aia["data"].shape

slices = (slice(0, 128), slice(0, 128))

if isinstance(wcs_out, BaseHighLevelWCS):
    low_level_wcs = SlicedLowLevelWCS(wcs_out.low_level_wcs, slices=slices)
else:
    low_level_wcs = SlicedLowLevelWCS(wcs_out, slices=slices)

wcs_out_sub = HighLevelWCSWrapper(low_level_wcs)

array = reproject_interp(
    input_data,
    wcs_out_sub,
    shape_out=shape_out,
    return_footprint=False,
)

fig = plt.figure()
ax = fig.add_subplot(1, 1, 1)
ax.imshow(array, origin="lower")
fig.savefig("chess.png")

gives:

[astrofrog]

This might be because I'm not supposed to change the forward transform like this, as the backward transform would also need to be changed?

[astrofrog]

Fixed now the file has been regenerated with https://github.com/astropy/reproject/pull/439