dask parallel need more check, to be simple, we set parallel to false…

… when open mfdataset; simplify caravan cache dataset

hydrodataset/caravan.py

@@ -619,12 +619,31 @@ def cache_attributes_xrdataset(self):
                 ds[var_name].attrs["units"] = converted_units[var_name]
         return ds
 
-    def cache_xrdataset(self):
-        """Save all attr data in a netcdf file in the cache directory, ts data are already nc format"""
+    def cache_xrdataset(self, checkregion="hysets"):
+        """
+        Save all attr data in a netcdf file in the cache directory,
+        ts data are already nc format
+
+        Parameters
+        ----------
+        checkregion : str, optional
+            as my experience, the dameged file is in hysets, by default "hysets"
+        """
         warnings.warn("Check you units of all variables")
         if self.region != "Global":
             raise ValueError("Only Global region is supported.")
-        pbar = tqdm(self.region_data_name, desc="Start Checking Data...")
+        ds_attr = self.cache_attributes_xrdataset()
+        ds_attr.to_netcdf(
+            os.path.join(
+                self.data_source_description["ATTR_DIR"],
+                "caravan_attributes.nc",
+            )
+        )
+        if checkregion is not None:
+            regions = [checkregion]
+        else:
+            regions = self.region_data_name
+        pbar = tqdm(regions, desc="Start Checking Data...")
         for region in pbar:
             pbar.set_description(f"Processing Region-{region}")
             # all files are too large to read in memory, hence we read them region by region
@@ -686,13 +705,6 @@ def cache_xrdataset(self):
                         os.path.dirname(file_path), f"{gage_id}.nc"
                     )
                     the_ncfile_data.to_netcdf(the_ncfile)
-        ds_attr = self.cache_attributes_xrdataset()
-        ds_attr.to_netcdf(
-            os.path.join(
-                self.data_source_description["ATTR_DIR"],
-                "caravan_attributes.nc",
-            )
-        )
 
     def read_attr_xrdataset(self, gage_id_lst=None, var_lst=None, **kwargs):
         # Define the path to the attributes file
@@ -727,8 +739,9 @@ def read_ts_xrdataset(self, gage_id_lst, t_range, var_lst, **kwargs):
             file_paths.append(file_path)
 
         # Open the dataset in a lazy manner using dask
+        parallel = kwargs.get("parallel", False)
         combined_ds = xr.open_mfdataset(
-            file_paths, combine="nested", concat_dim="basin", parallel=True
+            file_paths, combine="nested", concat_dim="basin", parallel=parallel
         )
 
         def extract_unit(variable_name, units_string):