Merge branch '408-remove-numpy-warbning' into 'release'

Resolve "Update après release 2.0.0 numpy"

See merge request 3d/PandoraBox/pandora!359

pandora/img_tools.py

@@ -685,7 +685,7 @@ def shift_right_img(img_right: xr.Dataset, subpix: int, band: str = None) -> Lis
             # For each index, shift the right image for subpixel precision 1/subpix*index
             data = zoom(selected_band, (1, (nx_ * subpix - (subpix - 1)) / float(nx_)), order=1)[:, ind::subpix]
             col = np.arange(
-                img_right.coords["col"][0] + shift * ind, img_right.coords["col"][-1], step=1
+                img_right.coords["col"].values[0] + shift * ind, img_right.coords["col"].values[-1], step=1
             )  # type: np.ndarray
             img_right_shift.append(
                 xr.Dataset(

pandora/matching_cost/matching_cost.py

@@ -330,7 +330,7 @@ def grid_estimation(
         :rtype: xarray.Dataset
         """
         # Get col dimension
-        c_col = img["im"].coords["col"]
+        c_col = img["im"].coords["col"].values
 
         # Get the index of the columns that should be computed
         if cfg and "ROI" in cfg:
@@ -569,7 +569,7 @@ def masks_dilatation(
 
             # Whatever the sub-pixel precision, only one sub-pixel mask is created,
             # since 0.5 shifted mask == 0.25 shifted mask
-            col_shift = np.arange(col[0] + 0.5, col[0] + nx_ - 1, step=1)  # type: np.ndarray
+            col_shift = np.arange(col.values[0] + 0.5, col.values[0] + nx_ - 1, step=1)  # type: np.ndarray
             dilatate_right_mask_shift = xr.DataArray(
                 dilatate_right_mask_shift, coords=[row, col_shift], dims=["row", "col"]
             )

setup.cfg

@@ -43,7 +43,7 @@ setup_requires =
 
 # Pandora packages dependencies
 install_requires =
-    numpy<2.0
+    numpy
     xarray>=0.13.0
     scipy
     rasterio

tests/test_matching_cost/test_matching_cost.py

@@ -323,7 +323,7 @@ def test_grid_estimation(self, default_input_roi, step, ground_truth_attrs):
 
         # Create ground truth for output of grid_estimation() function
         c_row = img_left["im"].coords["row"]
-        row = np.arange(c_row[0], c_row[-1] + 1)
+        row = np.arange(c_row.values[0], c_row.values[-1] + 1)
 
         ground_truth = xr.Dataset(
             {},
@@ -435,7 +435,7 @@ def test_grid_estimation_with_roi(self, default_input_roi, roi, step, ground_tru
 
         # Create ground truth for output of grid_estimation() function
         c_row = img_left["im"].coords["row"]
-        row = np.arange(c_row[0], c_row[-1] + 1)
+        row = np.arange(c_row.values[0], c_row.values[-1] + 1)
 
         ground_truth = xr.Dataset(
             {},