Fixes following numpy 2

pysteps/blending/steps.py

@@ -1098,7 +1098,7 @@ def worker(j):
                                 neginf=zerovalue,
                             )
                         # Put back the mask
-                        R_f_ip_recomp[domain_mask] = np.NaN
+                        R_f_ip_recomp[domain_mask] = np.nan
                         extrap_kwargs["displacement_prev"] = D[j]
                         R_f_ep_recomp_, D[j] = extrapolator(
                             R_f_ip_recomp,
@@ -1131,7 +1131,7 @@ def worker(j):
                                 neginf=np.nanmin(R_f_ip),
                             )
                         for i in range(n_cascade_levels):
-                            R_f_ep[i][temp_mask] = np.NaN
+                            R_f_ep[i][temp_mask] = np.nan
                         # B. Noise
                         Yn_ip_recomp = blending.utils.recompose_cascade(
                             combined_cascade=Yn_ip,
@@ -1182,7 +1182,7 @@ def worker(j):
                         extrap_kwargs_pb["displacement_prev"] = D_pb[j]
                         # Apply the domain mask to the extrapolation component
                         R_ = precip.copy()
-                        R_[domain_mask] = np.NaN
+                        R_[domain_mask] = np.nan
                         R_pm_ep_, D_pb[j] = extrapolator(
                             R_,
                             velocity_blended,

pysteps/feature/shitomasi.py

@@ -149,7 +149,7 @@ def detection(
 
     # convert to 8-bit
     input_image = np.ndarray.astype(input_image, "uint8")
-    mask = (-1 * mask + 1).astype("uint8")
+    mask = ~mask & 1
 
     params = dict(
         maxCorners=max_num_features if max_num_features is not None else max_corners,

pysteps/io/importers.py

@@ -810,15 +810,15 @@
     # the no data value. The precision of the data is two decimals (0.01 mm).
     if qty == "ACRR":
         precip = np.where(
-            precip_intermediate == 65535, np.NaN, precip_intermediate / 100.0
+            precip_intermediate == 65535, np.nan, precip_intermediate / 100.0
         )
 
     # In case reflectivities are imported, the no data value is 255. Values are
     # saved as integers. The reflectivities are not directly saved in dBZ, but
     # as: dBZ = 0.5 * pixel_value - 32.0 (this used to be 31.5).
     if qty == "DBZH":
         precip = np.where(
-            precip_intermediate == 255, np.NaN, precip_intermediate * 0.5 - 32.0
+            precip_intermediate == 255, np.nan, precip_intermediate * 0.5 - 32.0
         )
 
     if precip is None:
@@ -1366,9 +1366,9 @@
                                     mask = np.logical_and(~mask_u, ~mask_n)
                                     quality = np.empty(arr.shape)  # , dtype=float)
                                     quality[mask] = arr[mask] * gain + offset
-                                    quality[~mask] = (
-                                        np.nan
-                                    )  # a qui -----------------------------
+                                    quality[
+                                        ~mask
+                                    ] = np.nan  # a qui -----------------------------
 
     if precip is None:
         raise IOError("requested quantity %s not found" % qty)

pysteps/tests/test_io_nowcast_importers.py

@@ -18,6 +18,9 @@
     [(precip, metadata), (np.zeros_like(precip), metadata)],
 )
 def test_import_netcdf(precip, metadata, tmp_path):
+
+    pytest.importorskip("pyproj")
+
     field_shape = (precip.shape[1], precip.shape[2])
     startdate = metadata["timestamps"][-1]
     timestep = metadata["accutime"]

pysteps/tests/test_nowcasts_linda.py

@@ -144,6 +144,9 @@ def test_linda_wrong_inputs():
 
 def test_linda_callback(tmp_path):
     """Test LINDA callback functionality to export the output as a netcdf."""
+
+    pytest.importorskip("skimage")
+
     n_ens_members = 2
     n_timesteps = 3