chore: updating the attributes for the variables, coordinates and dat…

…aset (#93)

copernicusmarine/core_functions/describe.py

@@ -27,6 +27,7 @@ def describe_function(
     disable_progress_bar: bool,
     staging: bool,
 ) -> str:
+
     VersionVerifier.check_version_describe(staging)
     if staging:
         logger.warning(

copernicusmarine/download_functions/download_arco_series.py

@@ -97,7 +97,6 @@ def download_dataset(
     output_path = pathlib.Path(output_directory, filename)
     if not output_directory.is_dir():
         pathlib.Path.mkdir(output_directory, parents=True)
-
     if not force_download:
         logger.info(dataset)
         logger.info(

copernicusmarine/download_functions/subset_xarray.py

@@ -42,6 +42,36 @@
     "depth": ["depth", "deptht", "elevation"],
 }
 
+NETCDF_CONVENTION_VARIABLE_ATTRIBUTES = [
+    "standard_name",
+    "long_name",
+    "units",
+    "unit_long",
+    "valid_min",
+    "valid_max",
+]
+NETCDF_CONVENTION_COORDINATE_ATTRIBUTES = [
+    "standard_name",
+    "long_name",
+    "units",
+    "unit_long",
+    "axis",
+    "valid_min",
+    "valid_max",
+]
+NETCDF_CONVENTION_DATASET_ATTRIBUTES = [
+    "title",
+    "institution",
+    "source",
+    "history",
+    "references",
+    "comment",
+    "Conventions",
+    "producer",
+    "credit",
+    "contact",
+]
+
 
 def _dataset_custom_sel(
     dataset: xarray.Dataset,
@@ -280,6 +310,16 @@ def _get_variable_name_from_standard_name(
     return None
 
 
+def _update_variables_attributes(
+    dataset: xarray.Dataset, variables: List[str]
+) -> xarray.Dataset:
+    for variable in variables:
+        dataset[variable].attrs = _filter_attributes(
+            dataset[variable].attrs, NETCDF_CONVENTION_VARIABLE_ATTRIBUTES
+        )
+    return dataset
+
+
 def _variables_subset(
     dataset: xarray.Dataset, variables: List[str]
 ) -> xarray.Dataset:
@@ -298,17 +338,26 @@ def _variables_subset(
                 )
             else:
                 raise VariableDoesNotExistInTheDataset(variable)
-    return dataset[numpy.array(dataset_variables_filter)]
+    dataset = dataset[numpy.array(dataset_variables_filter)]
+    return _update_variables_attributes(dataset, dataset_variables_filter)
+
+
+def _filter_attributes(attributes: dict, attributes_to_keep: List[str]):
+    attributes_that_exist = set(attributes).intersection(attributes_to_keep)
+    return {key: attributes[key] for key in attributes_that_exist}
 
 
-def _update_dataset_coordinate_valid_minmax_attributes(
+def _update_dataset_coordinate_attributes(
     dataset: xarray.Dataset,
 ) -> xarray.Dataset:
     for coordinate_label in COORDINATES_LABEL:
         for coordinate_alias in COORDINATES_LABEL[coordinate_label]:
             if coordinate_alias in dataset.sizes:
                 coord = dataset[coordinate_alias]
                 attrs = coord.attrs
+                coordinate_attributes = (
+                    NETCDF_CONVENTION_COORDINATE_ATTRIBUTES.copy()
+                )
                 if "time" in coordinate_label:
                     min_time_dimension = coord.values.min()
                     max_time_dimension = coord.values.max()
@@ -319,13 +368,26 @@ def _update_dataset_coordinate_valid_minmax_attributes(
                     valid_max = convert_datetime64_to_netcdf_timestamp(
                         max_time_dimension, netcdf_unit
                     )
+                    attrs["standard_name"] = "time"
+                    attrs["long_name"] = "Time"
                     attrs["valid_min"] = valid_min
                     attrs["valid_max"] = valid_max
-                else:
+                    attrs["axis"] = "T"
+                    attrs["unit_long"] = (
+                        coord.encoding["units"].replace("_", " ").title()
+                    )
+                    coordinate_attributes.remove("units")
+                elif coordinate_label in ["latitude", "depth", "elevation"]:
                     attrs["valid_min"] = coord.values.min()
                     attrs["valid_max"] = coord.values.max()
+                elif coordinate_label == "longitude":
+                    coordinate_attributes.remove("valid_min")
+                    coordinate_attributes.remove("valid_max")
+                coord.attrs = _filter_attributes(attrs, coordinate_attributes)
 
-                coord.attrs = attrs
+    dataset.attrs = _filter_attributes(
+        dataset.attrs, NETCDF_CONVENTION_DATASET_ATTRIBUTES
+    )
 
     return dataset
 
@@ -351,7 +413,7 @@ def subset(
 
     dataset = _depth_subset(dataset, depth_parameters)
 
-    dataset = _update_dataset_coordinate_valid_minmax_attributes(dataset)
+    dataset = _update_dataset_coordinate_attributes(dataset)
 
     return dataset
 

tests/test_command_line_interface.py

@@ -557,8 +557,6 @@ def test_if_dataset_coordinate_valid_minmax_attributes_are_setted(
 
         assert dataset.latitude.attrs["valid_min"] >= 0
         assert dataset.latitude.attrs["valid_max"] <= 0.1
-        assert dataset.longitude.attrs["valid_min"] >= 0.2
-        assert dataset.longitude.attrs["valid_max"] <= 0.3
         assert dataset.depth.attrs["valid_min"] >= 0
         assert dataset.depth.attrs["valid_max"] <= 5
         assert dataset.time.attrs["valid_min"] == 648672
@@ -1093,26 +1091,24 @@ def then_I_have_correct_sign_for_depth_coordinates_values(
             assert dataset.elevation.min() >= -10
             assert dataset.elevation.max() <= 0
 
-    def then_I_have_correct_attribute_value(
-        self, output_path, dimention_name, attribute_value
-    ):
+    def then_I_have_correct_attribute_value(self, output_path, dimention_name):
         filepath = pathlib.Path(output_path, "data.zarr")
         dataset = xarray.open_dataset(filepath, engine="zarr")
-        assert dataset[dimention_name].attrs["positive"] == attribute_value
+        assert dataset[dimention_name].attrs["standard_name"] == dimention_name
 
     def test_conversion_between_elevation_and_depth(self, tmp_path):
         self.when_I_request_subset_dataset_with_zarr_service(tmp_path, True)
         self.then_I_have_correct_sign_for_depth_coordinates_values(
             tmp_path, "positive"
         )
-        self.then_I_have_correct_attribute_value(tmp_path, "depth", "down")
+        self.then_I_have_correct_attribute_value(tmp_path, "depth")
 
     def test_force_no_conversion_between_elevation_and_depth(self, tmp_path):
         self.when_I_request_subset_dataset_with_zarr_service(tmp_path, False)
         self.then_I_have_correct_sign_for_depth_coordinates_values(
             tmp_path, "negative"
         )
-        self.then_I_have_correct_attribute_value(tmp_path, "elevation", "up")
+        self.then_I_have_correct_attribute_value(tmp_path, "elevation")
 
     def when_I_run_copernicus_marine_command_using_no_directories_option(
         self, tmp_path, force_service: GetServiceToTest, output_directory=None

tests/test_python_interface.py

@@ -206,7 +206,6 @@ def test_subset_modify_attr_for_depth(self):
         dataset = open_dataset(
             dataset_id="cmems_mod_arc_phy_anfc_6km_detided_P1D-m"
         )
-        assert dataset.depth.attrs["positive"] == "down"
         assert dataset.depth.attrs["standard_name"] == "depth"
         assert dataset.depth.attrs["long_name"] == "Depth"
 
@@ -235,7 +234,10 @@ def test_subset_keeps_fillvalue_empty(self, tmp_path):
         assert "_FillValue" not in subsetdata.time.attrs
         assert "_FillValue" not in subsetdata.latitude.attrs
         assert "_FillValue" not in subsetdata.depth.attrs
-        assert "valid_max" in subsetdata.longitude.attrs
+        assert "valid_max" not in subsetdata.longitude.attrs
+        assert "valid_min" not in subsetdata.longitude.attrs
+        assert "valid_max" in subsetdata.latitude.attrs
+        assert "valid_max" in subsetdata.latitude.attrs
         assert subsetdata.time.attrs["calendar"] == "gregorian"
         assert subsetdata.time.attrs["units"] == "hours since 1950-01-01"
 
@@ -265,7 +267,6 @@ def test_subset_keeps_fillvalue_empty_w_compression(self, tmp_path):
         assert "_FillValue" not in subsetdata.time.attrs
         assert "_FillValue" not in subsetdata.latitude.attrs
         assert "_FillValue" not in subsetdata.depth.attrs
-        assert "valid_max" in subsetdata.longitude.attrs
         assert subsetdata.time.attrs["calendar"] == "gregorian"
         assert subsetdata.time.attrs["units"] == "hours since 1950-01-01"