Skip to content

Improve no overlap ref/dem behaviour in accuracy worflow #989

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
marinebcht wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Improve no overlap ref/dem behaviour in accuracy worflow #989

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
95a2559
add: improve no overlap behaviour
marinebcht Aug 7, 2026
ad1ca77
fix: structure
marinebcht Aug 7, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
84 changes: 47 additions & 37 deletions tests/test_workflows/test_accuracy.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,7 +25,6 @@
from pathlib import Path

import geoutils as gu
import numpy as np
import pandas as pd
import pytest

Expand Down Expand Up @@ -289,20 +288,16 @@ def test_run_prepare_datas(get_accuracy_inputs_test, tmp_path, config):
@pytest.mark.parametrize(
"config",
[
("reference_elev", ["reference_elev"]),
("reference_elev", ["to_be_aligned_elev"]),
("to_be_aligned_elev", ["reference_elev"]),
("to_be_aligned_elev", ["to_be_aligned_elev"]),
("reference_elev", ["reference_elev", "to_be_aligned_elev"]),

@marinebcht marinebcht Aug 7, 2026

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ex 5 and 6 : no overlap => moved to test_no_overlap()

("to_be_aligned_elev", ["reference_elev", "to_be_aligned_elev"]),
("reference_elev", "reference_elev"),
("reference_elev", "to_be_aligned_elev"),
("to_be_aligned_elev", "reference_elev"),
("to_be_aligned_elev", "to_be_aligned_elev"),
],
)
def test_prepare_datas(get_accuracy_inputs_test, tmp_path, config):
"""
Test preparation data with all sampling_grid values
"""
def test_prepare_datas_with_overlap(get_accuracy_inputs_test, tmp_path, config):
"""Test preparation data with all sampling_grid values when ref and tba do overlap."""

sampling_grid, dem_to_crop_list = config
sampling_grid, dem_to_crop = config
user_config = get_accuracy_inputs_test

# Save path before crop(s)
Expand All @@ -325,11 +320,10 @@ def test_prepare_datas(get_accuracy_inputs_test, tmp_path, config):
crop["to_be_aligned_elev"] = (int(ncols / 2) + 1, int(nrows / 2) + 1, ncols, nrows)

# Crop dems(s) and update config
for dem_to_crop in dem_to_crop_list:
dem = xdem.DEM(user_config["inputs"][dem_to_crop]["path_to_elev"])
dem_crop = dem.icrop(crop[dem_to_crop])
dem_crop.to_file(Path(tmp_path / (dem_to_crop + "_crop.tif")))
user_config["inputs"][dem_to_crop]["path_to_elev"] = Path(tmp_path / (dem_to_crop + "_crop.tif")).as_posix()
dem = xdem.DEM(user_config["inputs"][dem_to_crop]["path_to_elev"])
dem_crop = dem.icrop(crop[dem_to_crop])
dem_crop.to_file(Path(tmp_path / (dem_to_crop + "_crop.tif")))
user_config["inputs"][dem_to_crop]["path_to_elev"] = Path(tmp_path / (dem_to_crop + "_crop.tif")).as_posix()

workflows = Accuracy(user_config)
workflows.run()
Expand All @@ -356,36 +350,52 @@ def test_prepare_datas(get_accuracy_inputs_test, tmp_path, config):
).get_stats("mean")

if sampling_grid == "reference_elev":

# If reference_elev is cropped or not
dem_ref_ref = xdem.DEM(original_ref_path)
if "reference_elev" in dem_to_crop_list:
# If reference_elev is cropped (input)
if "reference_elev" == dem_to_crop:
dem_ref_ref = dem_ref_ref.icrop(crop["reference_elev"])
assert dem_ref_ref.get_stats("mean") == reference_elev_reprojected_mean

# If intersection between ref and tba is not null
if len(dem_to_crop_list) == 1:
assert xdem.DEM(original_tba_path).icrop(crop[dem_to_crop_list[0]]).get_stats("mean") == pytest.approx(
to_be_aligned_elev_reprojected_mean
)
else:
assert np.isnan(to_be_aligned_elev_reprojected_mean)
assert xdem.DEM(original_tba_path).icrop(crop[dem_to_crop]).get_stats("mean") == pytest.approx(
to_be_aligned_elev_reprojected_mean
)

else:

# If to_be_aligned_elev is cropped or not
dem_tba_ref = xdem.DEM(original_tba_path)
if "to_be_aligned_elev" in dem_to_crop_list:
# If to_be_aligned_elev is cropped (input)
if "to_be_aligned_elev" == dem_to_crop:
dem_tba_ref = dem_tba_ref.icrop(crop["to_be_aligned_elev"])
assert dem_tba_ref.get_stats("mean") == to_be_aligned_elev_reprojected_mean

# If intersection between ref and tba is not null
if len(dem_to_crop_list) == 1:
assert xdem.DEM(original_ref_path).icrop(crop[dem_to_crop_list[0]]).get_stats("mean") == pytest.approx(
reference_elev_reprojected_mean
)
else:
assert np.isnan(reference_elev_reprojected_mean)
assert xdem.DEM(original_ref_path).icrop(crop[dem_to_crop]).get_stats("mean") == pytest.approx(
reference_elev_reprojected_mean
)


@pytest.mark.parametrize("process", [True, False])
@pytest.mark.parametrize("sampling_grid", ["reference_elev", "to_be_aligned_elev"])
def test_no_overlap(get_accuracy_inputs_test, tmp_path, process, sampling_grid):
"""Test coreg/no coreg processes when ref and tba do not overlap."""
user_config = get_accuracy_inputs_test
user_config["inputs"]["sampling_grid"] = sampling_grid
if not process:
user_config["coregistration"] = {"process": False}

ref_dem = xdem.DEM(user_config["inputs"]["reference_elev"]["path_to_elev"])
nrows, ncols = ref_dem.shape
ref_dem = ref_dem.icrop((ncols - int(ncols / 2), nrows - int(nrows / 2), ncols, nrows))
ref_dem.to_file(tmp_path / "ref_cropped.tif")
user_config["inputs"]["reference_elev"]["path_to_elev"] = str(tmp_path / "ref_cropped.tif")

tba_dem = xdem.DEM(user_config["inputs"]["to_be_aligned_elev"]["path_to_elev"])
nrows, ncols = tba_dem.shape
tba_dem = tba_dem.icrop((0, 0, ncols - int(ncols / 2), nrows - int(nrows / 2)))
tba_dem.to_file(tmp_path / "tba_cropped.tif")
user_config["inputs"]["to_be_aligned_elev"]["path_to_elev"] = str(tmp_path / "tba_cropped.tif")

workflows = Accuracy(user_config)
with pytest.raises(ValueError, match="Reference and To-be-align elevations do not overlap."):

@rhugonnet rhugonnet Aug 13, 2026 •
edited
Loading

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's try to draft a more helpful message for users with more details and potential solutions, maybe something along those lines:
"Reference and to-be-aligned elevation datasets do not overlap horizontally. Their coregistration method '{coreg_method}' relies on elevation differencing, and thus produces only NaNs. To resolve this: If a large initial misalignment is expected, use a first coregistration method not requiring initial overlap such as ICP or CPD. If no large misalignment was expected, check the georeferencing of your data and re-set it manually with set_crs or set_transform."

workflows.run()


@pytest.mark.parametrize(
Expand Down
3 changes: 3 additions & 0 deletions xdem/workflows/accuracy.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -200,6 +200,9 @@ def _prepare_datas(self, vmin: float, vmax: float) -> None:
elif sampling_grid == "to_be_aligned_elev":
self.reference_elev = self.reference_elev.reproject(self.to_be_aligned_elev, silent=True)

if not self.reference_elev.get_stats("validcount") or not self.to_be_aligned_elev.get_stats("validcount"):
raise ValueError("Reference and To-be-align elevations do not overlap.")

# Intersection
logging.info("Computing intersection")
coord_intersection = self.reference_elev.intersection(self.to_be_aligned_elev)
Expand Down
Loading