Update Docs Issue 560

CONTRIBUTING.md

@@ -33,7 +33,7 @@ Please [make the pull request a draft](https://github.blog/2019-02-14-introducin
 
 Currently, we have two regular reviewers of pull requests:
 
-  * Kevin Santana ([kevinsantana11](https://github.com/kevinsantana11))
+  * Kevin Shuman ([KevinShuman](https://github.com/KevinShuman))
   * Shane Elipot ([selipot](https://github.com/selipot))
 
 You can request a review from one of us or just comment in GitHub that you want a review and we'll see it. Only one review is required to be allowed to merge a pull request. We'll work with you to get it into shape.
@@ -228,7 +228,7 @@ make html
 
 Currently, only one person can deploy releases:
 
-  * Kevin Santana ([kevinsantana11](https://github.com/kevinsantana11))
+  * Kevin Shuman ([KevinShuman](https://github.com/KevinShuman))
 
 If you need your merged pull request to be deployed in a release, just ask!
 

README.md

@@ -141,7 +141,6 @@ cd clouddrift/
 2. Create an environment as specified in the yml file with the required library dependencies:
 ```bash
 conda env create -f environment.yml # creates a new env with the dependencies
-conda env update -f environment.yml # install dependencies in current environment
 ```
 
 2a. Make sure you **created** the environment by activating it:

clouddrift/adapters/gdp/gdpsource.py

@@ -390,7 +390,7 @@
 
     # Transform the initial dataframe filtering out rows with really anomolous values
     # examples include: years in the future, years way in the past before GDP program, etc...
-    preremove_df_chunk = df_chunk.assign(obs_index=range(start_idx, end_idx))
+    preremove_df_chunk = df_chunk.assign(obs_index=np.arange(start_idx, end_idx))
     df_chunk = _apply_remove(
         preremove_df_chunk,
         filters=[

clouddrift/adapters/subsurface_floats.py

@@ -61,7 +61,8 @@
 
     metadata = {}
     for var in meta_variables:
-        metadata[var] = np.array([v.flatten()[0] for v in source_data[var].flatten()])
+        arrs = _to_dense_flatten(source_data[str(var)])
+        metadata[var] = np.array([_flatten_array(v)[0] for v in arrs])
 
     # bring the expList to the "traj" dimension
     _, float_per_exp = np.unique(metadata["indexExp"], return_counts=True)
@@ -71,11 +72,12 @@
     data_variables = ["dtnum", "lon", "lat", "p", "t", "u", "v"]
     data = {}
     for var in data_variables:
-        data[var] = np.concatenate([v.flatten() for v in source_data[var].flatten()])
+        arrs = _to_dense_flatten(source_data[str(var)])
+        data[var] = np.concatenate([_flatten_array(v) for v in arrs])
 
     # create rowsize variable
-    rowsize = np.array([len(v) for v in source_data["dtnum"].flatten()])
-    assert np.sum(rowsize) == len(data["dtnum"])
+    arrs = _to_dense_flatten(source_data["dtnum"])
+    rowsize = np.array([len(_flatten_array(v)) for v in arrs])
 
     # Unix epoch start (1970-01-01)
     origin_datenum = 719529
@@ -197,3 +199,17 @@
     ds = ds.set_coords(["time", "id"])
 
     return ds
+
+
+def _flatten_array(arr):
+    # Convert sparse to dense if needed, then flatten
+    if hasattr(arr, "toarray"):
+        arr = arr.toarray()
+    return np.array(arr).flatten()
+
+
+def _to_dense_flatten(arr):
+    """Convert a possibly sparse array to dense and flatten it."""
+    if hasattr(arr, "toarray"):
+        arr = arr.toarray()
+    return np.array(arr).flatten()

clouddrift/transfer.py

@@ -4,6 +4,7 @@
 """
 
 import numpy as np
+from numpy import floating as _Floating
 from numpy.lib.scimath import sqrt
 from scipy.special import factorial, iv, kv  # type: ignore
 
@@ -63,7 +64,11 @@
     boundary_condition: str = "no-slip",
     method: str = "lilly",
     density: float = 1025.0,
-) -> tuple[np.ndarray, np.ndarray, np.ndarray]:
+) -> tuple[
+    float | _Floating | np.ndarray,
+    float | _Floating | np.ndarray,
+    float | _Floating | np.ndarray,
+]:
     """
     Compute the transfer function from wind stress to oceanic velocity based on the physically-based
     models of Elipot and Gille (2009) and Lilly and Elipot (2021).
@@ -677,57 +682,61 @@
 def _bessels_freeslip(
     xiz: float | np.ndarray,
     xih: float | np.ndarray,
-    xi0: float | np.ndarray | None | None = None,
-) -> tuple[
-    float | np.ndarray,
-    float | np.ndarray,
-    float | np.ndarray,
-    float | np.ndarray,
-    float | np.ndarray,
-    float | np.ndarray,
-]:
+    xi0: float | np.ndarray | None = None,
+) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray]:
     """
     Compute the Bessel functions for the free-slip boundary condition for the xsi(z), xsi(h), and xsi(0) functions.
     """
-    k0z = kv(0, xiz)
-    i0z = iv(0, xiz)
-    k1h = kv(1, xih)
-    i1h = iv(0, xih)
+    # Convert inputs to numpy arrays
+    xiz = np.asarray(xiz)
+    xih = np.asarray(xih)
+
+    # Ensure all outputs are numpy arrays
+    k0z = np.asarray(kv(0, xiz))
+    i0z = np.asarray(iv(0, xiz))
+    k1h = np.asarray(kv(1, xih))
+    i1h = np.asarray(iv(0, xih))
 
     if xi0 is not None:
-        k10 = kv(1, xi0)
-        i10 = iv(1, xi0)
-        return k0z, i0z, k1h, i1h, k10, i10
+        xi0 = np.asarray(xi0)
+        k10 = np.asarray(kv(1, xi0))
+        i10 = np.asarray(iv(1, xi0))
     else:
-        return k0z, i0z, k1h, i1h, np.nan, np.nan
+        # Create nan values as numpy arrays with same shape as k0z
+        k10 = np.full_like(k0z, np.nan)
+        i10 = np.full_like(k0z, np.nan)
+
+    return k0z, i0z, k1h, i1h, k10, i10
 
 
 def _bessels_noslip(
     xiz: float | np.ndarray,
     xih: float | np.ndarray,
     xi0: float | np.ndarray | None = None,
-) -> tuple[
-    float | np.ndarray,
-    float | np.ndarray,
-    float | np.ndarray,
-    float | np.ndarray,
-    float | np.ndarray,
-    float | np.ndarray,
-]:
+) -> tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray, np.ndarray]:
     """
     Compute the Bessel functions for the no-slip boundary condition for the xsi(z), xsi(h), and xsi(0) functions.
     """
-    k0z = kv(0, xiz)
-    i0z = iv(0, xiz)
-    k0h = kv(0, xih)
-    i0h = iv(0, xih)
+    # Convert inputs to numpy arrays
+    xiz = np.asarray(xiz)
+    xih = np.asarray(xih)
+
+    # Ensure all outputs are numpy arrays
+    k0z = np.asarray(kv(0, xiz))
+    i0z = np.asarray(iv(0, xiz))
+    k0h = np.asarray(kv(0, xih))
+    i0h = np.asarray(iv(0, xih))
 
     if xi0 is not None:
-        k10 = kv(1, xi0)
-        i10 = iv(1, xi0)
-        return k0z, i0z, k0h, i0h, k10, i10
+        xi0 = np.asarray(xi0)
+        k10 = np.asarray(kv(1, xi0))
+        i10 = np.asarray(iv(1, xi0))
     else:
-        return k0z, i0z, k0h, i0h, np.nan, np.nan
+        # Create nan values as numpy arrays with same shape as k0z
+        k10 = np.full_like(k0z, np.nan)
+        i10 = np.full_like(k0z, np.nan)
+
+    return k0z, i0z, k0h, i0h, k10, i10
 
 
 def _besseltildes_noslip(