Change "idw" to "kriging" in spatial_interpolation_kernel parameters

src/spikeinterface/preprocessing/motion.py

@@ -69,10 +69,7 @@
             weight_with_amplitude=False,
         ),
         "interpolate_motion_kwargs": dict(
-            direction=1,
-            border_mode="remove_channels",
-            spatial_interpolation_method="idw",
-            num_closest=3,
+            direction=1, border_mode="remove_channels", spatial_interpolation_method="kriging", sigma_um=20.0, p=2
         ),
     },
     "nonrigid_fast_and_accurate": {
@@ -131,10 +128,7 @@
             weight_with_amplitude=False,
         ),
         "interpolate_motion_kwargs": dict(
-            direction=1,
-            border_mode="remove_channels",
-            spatial_interpolation_method="idw",
-            num_closest=3,
+            direction=1, border_mode="remove_channels", spatial_interpolation_method="kriging", sigma_um=20.0, p=2
         ),
     },
     # This preset is a super fast rigid estimation with center of mass
@@ -159,10 +153,7 @@
             rigid=True,
         ),
         "interpolate_motion_kwargs": dict(
-            direction=1,
-            border_mode="remove_channels",
-            spatial_interpolation_method="idw",
-            num_closest=3,
+            direction=1, border_mode="remove_channels", spatial_interpolation_method="kriging", sigma_um=20.0, p=2
         ),
     },
     # This preset try to mimic kilosort2.5 motion estimator
@@ -196,11 +187,7 @@
             win_shape="rect",
         ),
         "interpolate_motion_kwargs": dict(
-            direction=1,
-            border_mode="force_extrapolate",
-            spatial_interpolation_method="kriging",
-            sigma_um=[20.0, 30],
-            p=1,
+            direction=1, border_mode="force_extrapolate", spatial_interpolation_method="kriging", sigma_um=20.0, p=2
         ),
     },
     # empty preset

src/spikeinterface/preprocessing/preprocessing_tools.py

@@ -9,7 +9,7 @@ def get_spatial_interpolation_kernel(
     method="kriging",
     sigma_um=20.0,
     p=1,
-    num_closest=3,
+    num_closest=4,
     sparse_thresh=None,
     dtype="float32",
     force_extrapolate=False,
@@ -40,7 +40,7 @@ def get_spatial_interpolation_kernel(
         Used in the "kriging" formula
     sparse_thresh: None or float, default: None
         If not None for "kriging" force small value to be zeros to get a sparse matrix.
-    num_closest: int, default: 3
+    num_closest: int, default: 4
         Used for "idw"
     force_extrapolate: bool, default: False
         How to handle when target location are outside source location.

src/spikeinterface/preprocessing/tests/test_preprocessing_tools.py

@@ -13,6 +13,7 @@ def test_get_spatial_interpolation_kernel():
     target_location[:, 1] = np.linspace(0, 130, 4)
 
     for method in ("kriging", "idw", "nearest"):
+
         kernel = get_spatial_interpolation_kernel(
             source_location, target_location, method=method, force_extrapolate=False
         )
@@ -23,6 +24,13 @@ def test_get_spatial_interpolation_kernel():
         # check sum sum close 1
         assert np.allclose(np.sum(kernel[:, 1:], axis=0), 1.0)
 
+    #     import matplotlib.pyplot as plt
+    #     fig, ax = plt.subplots()
+    #     im = ax.imshow(kernel)
+    #     ax.set_title(method)
+    #     fig.colorbar(im)
+    # plt.show()
+
 
 if __name__ == "__main__":
     test_get_spatial_interpolation_kernel()