Add a temporary workaround to rename nv bounds dim

Author: tomvothecoder

e3sm_diags/driver/arm_diags_driver.py

@@ -18,6 +18,7 @@
 from e3sm_diags.logger import _setup_child_logger
 from e3sm_diags.plot import arm_diags_plot
 
+
 if TYPE_CHECKING:
     from e3sm_diags.parameter.arm_diags_parameter import ARMDiagsParameter
 
@@ -278,6 +279,11 @@ def _run_diag_annual_cycle(parameter: ARMDiagsParameter) -> ARMDiagsParameter:
                     vars_funcs = _get_vars_funcs_for_derived_var(ds_ref, var)
                     target_var = list(vars_funcs.keys())[0][0]
 
+                    # NOTE: The bounds dimension can be "nv", which is not
+                    # currently recognized as a valid bounds dimension
+                    # by xcdat. We rename it to "bnds" to make it compatible.
+                    ds_ref = _rename_bounds_dim(ds_ref)
+
                     ds_ref_climo = ds_ref.temporal.climatology(target_var, "month")
                     da_ref = vars_funcs[(target_var,)](ds_ref_climo[target_var]).rename(
                         var
@@ -530,3 +536,33 @@ def _save_metrics_to_json(parameter: ARMDiagsParameter, metrics_dict: Dict[str,
         json.dump(metrics_dict, outfile)
 
     logger.info(f"Metrics saved in: {abs_path}")
+
+
+def _rename_bounds_dim(ds: xr.Dataset) -> xr.Dataset:
+    """
+    Renames the bounds dimension "nv" to "bnds" in the given xarray.Dataset for
+    xCDAT compatibility.
+
+    This is a temporary workaround to ensure compatibility with xCDAT's bounds
+    handling. The bounds dimension "nv" is commonly used in datasets to
+    represent the number of vertices in a polygon, but xCDAT expects the
+    bounds dimension to be in `xcdat.bounds.VALID_BOUNDS_DIMS`. This function
+    renames "nv" to "bnds" to align with xCDAT's expectations.
+
+    Parameters
+    ----------
+    ds : xr.Dataset
+        The input xarray.Dataset which may contain a bounds dimension named "nv".
+
+    Returns
+    -------
+    xr.Dataset
+        A new xarray.Dataset with the "nv" dimension renamed to "bnds" if it
+        existed; otherwise, the original dataset copy.
+    """
+    ds_new = ds.copy()
+
+    if "nv" in ds_new.dims:
+        ds_new = ds_new.rename({"nv": "bnds"})
+
+    return ds_new