Moved model-specific parameters from config to class

config_example.yml

 model_name: esrm20
 exposure_format: esrm20  # Only supported value for now
 metadata_filepath: full_path_to_xlsx_file_with_model_metadata
-occupancy_cases:  # Occupancy cases to be processed
-  residential:
-    sheet_name: RES  # Sheet name in metadata_filepath
-    data_units_types_field: fieldnameRes  # Field name in sheet_name
-  commercial:
-    sheet_name: COM  # Sheet name in metadata_filepath
-    data_units_types_field: fieldnameCom  # Field name in sheet_name
-  industrial:
-    sheet_name: IND  # Sheet name in metadata_filepath
-    data_units_types_field: fieldnameInd  # Field name in sheet_name

gdeimporter/aggregatedexposuremodel.py

@@ -33,21 +33,25 @@ class AggregatedExposureModel(abc.ABC):
             Name of the input aggregated model.
         self.exposure_format (str):
             Format of the input aggregated model. Currently supported values: "esrm20".
-        self.occupancy_cases (list of str):
-            List of names of the occupancy cases for which the input aggregated exposure model
-            is defined.
+        self.occupancy_cases (dict):
+            Dictionary in which each first level key corresponds to an occupancy case (e.g.
+            "residential", "commercial", "industrial") for which the input aggregated exposure
+            model is defined. The subkeys will point at specific mappings for each sub-class.
         self.exposure_entities (dictionary of ExposureEntity):
             Dictionary of instances of ExposureEntity objects, each of which represent an
             exposure entity where the input aggregated exposure model is defined. See
             attributes in description of ExposureEntity. The keys of the dictionary are the
             names of the corresponding exposure entities.
+        self.filename_pattern (str):
+            Pattern of the names of the files that define the input aggregated exposure model.
     """
 
     def __init__(self, configuration):
         self.model_name = configuration.model_name
         self.exposure_format = configuration.exposure_format
-        self.occupancy_cases = list(configuration.occupancy_cases.keys())
-        self.exposure_entities = self.retrieve_exposure_entities(configuration)
+        self.occupancy_cases = None
+        self.exposure_entities = None
+        self.filename_pattern = None
 
     def retrieve_exposure_entities(self, configuration):
         """This function retrieves the exposure entities for which an input aggregated exposure
@@ -79,8 +83,56 @@ class ExposureModelESRM20(AggregatedExposureModel):
     model.
 
     See details in https://git.gfz-potsdam.de/dynamicexposure/datasources/-/tree/master/esrm20.
+
+    Attributes:
+        self.occupancy_cases (dict):
+            Dictionary in which each first level key corresponds to an occupancy case (e.g.
+            "residential", "commercial", "industrial") for which ESRM20 is defined. Each first
+            level key contains three sub-keys: "short", "sheet_name" and
+            "data_units_types_field":
+                short (str):
+                    Name of the occupancy case as found in the ESRM20 model files.
+                sheet_name (str):
+                    Name of the sheet in the meatadata file of the input aggregated exposure
+                    model from which info on this occupancy case can be retrieved.
+                data_units_types_field (str):
+                    Name of the field in sheet_name from which to retrieve information on the
+                    types of data_units.
+        self.filename_pattern (str):
+            Pattern of the names of the ESRM20 CSV files.
+        self.exposure_entities (dictionary of ExposureEntity):
+            Dictionary of instances of ExposureEntity objects, each of which represent an
+            exposure entity where the input aggregated exposure model is defined. See
+            attributes in description of ExposureEntity. The keys of the dictionary are the
+            names of the corresponding exposure entities.
     """
 
+    def __init__(self, configuration):
+        super().__init__(configuration)
+        self.occupancy_cases = {
+            "residential": {
+                "short": "Res",
+                "sheet_name": "RES",
+                "data_units_types_field": "Admin level resolution/aggregation",
+            },
+            "commercial": {
+                "short": "Com",
+                "sheet_name": "COM",
+                "data_units_types_field": "Admin level resolution/aggregation",
+            },
+            "industrial": {
+                "short": "Ind",
+                "sheet_name": "IND",
+                "data_units_types_field": "Resolution",
+            },
+        }
+        self.filename_pattern = {
+            "filename": "Exposure_Model_%s_%s.csv",
+            "first": "entity_name",
+            "second": "occupancy_short",
+        }
+        self.exposure_entities = self.retrieve_exposure_entities(configuration)
+
     def retrieve_exposure_entities(self, configuration):
         """This function retrieves the exposure entities of the ESRM20 model from the
         corresponding .xlsx metadata file.
@@ -94,17 +146,6 @@ class ExposureModelESRM20(AggregatedExposureModel):
                     metadata_filepath (string):
                         Full file path to the .xlsx file that contains metadata on the input
                         aggregated exposure model.
-                    occupancy_cases (dictionary):
-                        Dictionary in which each first level key corresponds to an occupancy
-                        case (e.g. "residential", "commercial", "industrial"). Each first level
-                        key contains two sub-keys: "sheet_name" and "data_units_types_field":
-                            sheet_name (str):
-                                Name of the sheet in the meatadata file of the input aggregated
-                                exposure model from which info on this occupancy case can be
-                                retrieved. E.g.: "RES", "COM", "IND".
-                            data_units_types_field (str):
-                                Name of the field in sheet_name from which to retrieve
-                                information on the types of data_units. E.g.: "Resolution".
 
         Returns:
             exposure_entities (dictionary of ExposureEntity):
@@ -124,12 +165,12 @@ class ExposureModelESRM20(AggregatedExposureModel):
         # Needs to go by occupancy case because the names and properties of the exposure
         # entities can only be read from the metadata file for a sheet that is associated
         # with a particular occupancy case:
-        for case in configuration.occupancy_cases.keys():
+        for case in self.occupancy_cases.keys():
 
             # Read the file (errors will be handled by pandas)
             metadata = pandas.read_excel(
                 configuration.metadata_filepath,
-                sheet_name=configuration.occupancy_cases[case]["sheet_name"],
+                sheet_name=self.occupancy_cases[case]["sheet_name"],
                 header=None,  # Otherwise we cannot handle repeated column names properly
                 index_col=0,  # Use first column as index
             )
@@ -160,7 +201,7 @@ class ExposureModelESRM20(AggregatedExposureModel):
 
             # Retrieve the row from which the types of data units can be interpreted
             data_units_types_row = metadata.loc[
-                configuration.occupancy_cases[case]["data_units_types_field"], :
+                self.occupancy_cases[case]["data_units_types_field"], :
             ]
             if len(data_units_types_row.shape) > 1:  # This should not occur
                 logger.critical(

gdeimporter/configuration.py

@@ -33,20 +33,9 @@ class Configuration:
         self.metadata_filepath (str):
             Full file path to the .xlsx file that contains metadata on the input aggregated
             exposure model.
-        self.occupancy_cases (dictionary):
-            Dictionary in which each first level key corresponds to an occupancy case (e.g.
-            "residential", "commercial", "industrial"). Each first level key contains two
-            sub-keys: "sheet_name" and "data_units_types_field":
-                sheet_name (str):
-                    Name of the sheet in the meatadata file of the input aggregated exposure
-                    model from which info on this occupancy case can be retrieved. E.g.: "RES",
-                    "COM", "IND".
-                data_units_types_field (str):
-                    Name of the field in sheet_name from which to retrieve information on the
-                    types of data_units. E.g.: "Resolution".
     """
 
-    REQUIRES = ["metadata_filepath", "occupancy_cases"]
+    REQUIRES = ["metadata_filepath"]
 
     def __init__(self, filepath):
         """
@@ -61,9 +50,6 @@ class Configuration:
 
         self.exposure_format = self._assign_parameter(config, "exposure_format")
         self.metadata_filepath = self._assign_parameter(config, "metadata_filepath")
-        self.occupancy_cases = self._assign_hierarchical_parameters(
-            config, "occupancy_cases", ["sheet_name", "data_units_types_field"]
-        )
 
         # Terminate if critical parameters are missing (not all parameters are critical)
         for key_parameter in self.REQUIRES:
@@ -122,51 +108,3 @@ class Configuration:
             assigned_parameter = None
 
         return assigned_parameter
-
-    def _assign_hierarchical_parameters(self, config, input_parameter, requested_nested):
-        """This function searches for the key input_parameter in the dictionary config, and for
-        each of the elements of requested_nested as keys of config[input_parameter].
-
-        If input_parameter is not a key of config, the output is None.
-
-        If input_parameter is a key of config, but one of the elements of requested_nested is
-        not a key of config[input_parameter]
-
-        Args:
-            config (dictionary):
-                The configuration file read as a dictionary. It may be an empty dictionary.
-            input_parameter (str):
-                Name of the desired parameter, to be searched for as a primary key of config.
-            requested_nested (list of str):
-                List of the names of the desired nested parameters, to be searched for as keys
-                of config[input_parameter].
-
-        Returns:
-            assigned_parameter (dictionary or None):
-                The content of config[input_parameter], if input_parameter is a key of config
-                and all elements of requested_nested are keys of config[input_parameter], or
-                None otherwise.
-        """
-
-        assigned_parameter = self._assign_parameter(config, input_parameter)
-
-        if assigned_parameter is None:
-            return None
-
-        if not isinstance(assigned_parameter, dict):
-            return None
-
-        sub_parameters_missing = False
-        for case in assigned_parameter.keys():
-            for requested_parameter in requested_nested:
-                if requested_parameter not in assigned_parameter[case].keys():
-                    logger.critical(
-                        "ERROR instantiating Configuration: occupancy case '%s' does not "
-                        "contain a '%s' parameter" % (case, requested_parameter)
-                    )
-                    sub_parameters_missing = True
-
-        if sub_parameters_missing is True:
-            assigned_parameter = None
-
-        return assigned_parameter

tests/data/config_for_testing_missing_01.yml

-model_name: esrm20
-exposure_format: esrm20
-metadata_filepath: /some/path/metadata.xlsx
-occupancy_cases:
-  residential:
-    sheet_names: RES
-    data_units_types_field: Admin level resolution/aggregation
-  commercial:
-    sheet_name: COM
-    data_units_types_field: Admin level resolution/aggregation
-  industrial:
-    sheet_name: IND
-    data_units_types_field: Resolution

tests/data/config_for_testing_missing_02.yml

-model_name: esrm20
-exposure_format: esrm20
-metadata_filepath: /some/path/metadata.xlsx
-occupancy_cases:
-  residential:
-    sheet_name: RES
-    data_units_types_field: Admin level resolution/aggregation
-  commercial:
-    sheet_name: COM
-    another_name_that_is_wrong: Admin level resolution/aggregation
-  industrial:
-    sheet_name: IND
-    data_units_types_field: Resolution

tests/data/config_for_testing_missing_03.yml

-model_name: esrm20
-exposure_format: esrm20
-metadata_filepath: /home/cnievas/Documents/GFZ_local/SERA/From_EFEHR_GitLab/TESTING/esrm20_exposure/sources/European_Exposure_Model_Data_Inputs_Sources.xlsx
-occupancy_cases: residential

tests/test_configuration.py

@@ -29,21 +29,6 @@ def test_Configuration():
     assert returned_config.model_name == "esrm20"
     assert returned_config.exposure_format == "esrm20"
     assert returned_config.metadata_filepath == "/some/path/metadata.xlsx"
-    assert len(returned_config.occupancy_cases.keys()) == 3
-    assert returned_config.occupancy_cases["residential"]["sheet_name"] == "RES"
-    assert (
-        returned_config.occupancy_cases["residential"]["data_units_types_field"]
-        == "Admin level resolution/aggregation"
-    )
-    assert returned_config.occupancy_cases["commercial"]["sheet_name"] == "COM"
-    assert (
-        returned_config.occupancy_cases["commercial"]["data_units_types_field"]
-        == "Admin level resolution/aggregation"
-    )
-    assert returned_config.occupancy_cases["industrial"]["sheet_name"] == "IND"
-    assert (
-        returned_config.occupancy_cases["industrial"]["data_units_types_field"] == "Resolution"
-    )
 
     # Test case in which the file is not found
     with pytest.raises(OSError) as excinfo:
@@ -51,24 +36,3 @@ def test_Configuration():
             os.path.join(os.path.dirname(__file__), "data", "doesnotexist.yml")
         )
     assert "OSError" in str(excinfo.type)
-
-    # Test case in which the parameter "sheet_name" is missing
-    with pytest.raises(OSError) as excinfo:
-        returned_config = Configuration(
-            os.path.join(os.path.dirname(__file__), "data", "config_for_testing_missing_01.yml")
-        )
-    assert "OSError" in str(excinfo.type)
-
-    # Test case in which the parameter "data_units_types_field" is missing
-    with pytest.raises(OSError) as excinfo:
-        returned_config = Configuration(
-            os.path.join(os.path.dirname(__file__), "data", "config_for_testing_missing_02.yml")
-        )
-    assert "OSError" in str(excinfo.type)
-
-    # Test case in which the parameter "occupancy_cases" is not hierarchical
-    with pytest.raises(OSError) as excinfo:
-        returned_config = Configuration(
-            os.path.join(os.path.dirname(__file__), "data", "config_for_testing_missing_03.yml")
-        )
-    assert "OSError" in str(excinfo.type)