implemented unit check for extensions char

pymrio/core/mriosystem.py

@@ -1934,7 +1934,10 @@ def characterize(
 
     def convert(self, df_map, extension_name, 
                 agg_func="sum", 
-                drop_not_bridged_index=True):
+                drop_not_bridged_index=True,
+                unit_column_orig="unit_orig",
+                unit_column_new="unit_new",
+                ignore_columns=None):
         """ Apply the convert function to all dataframes in the extension
 
         Parameters
@@ -2001,27 +2004,55 @@ def convert(self, df_map, extension_name,
             make a bridge column for the ones to be dropped and map all to the same name.
             Then drop this index level after the conversion.
 
+        unit_column_orig : str, optional
+            Name of the column in df_map with the original unit.
+            This will be used to check if the unit matches the original unit in the extension.
+            Default is "unit_orig", if None, no check is performed.
 
+        unit_column_new : str, optional
+            Name of the column in df_map with the new unit to be assigned to the new extension.
+            Default is "unit_new", if None same unit as in df_orig TODO EXPLAIN BETTER, THINK WARNING
+
+        ignore_columns : list, optional
+            List of column names in df_map which should be ignored.
+            These could be columns with additional information, etc.
+            The unit columns given in unit_column_orig and unit_column_new 
+            are ignored by default.
+
+
+        TODO: remove after explain
         Extension for extensions:
         extension ... extension name
         unit_orig ... the original unit (optional, for double check with the unit)
         unit_new ... the new unit to be set for the extension
 
         """
-        # unit, unit_new, unit_orig
-        if "unit_orig" in df_map.columns:
-            # check if the units in the extensions confirm the units
-            pass
-        if "unit" in df_map.columns or "unit_new" in df_map.columns:
-            pass
-            # set the new unit column
-        else:
-            pass
-            # set unit to "undef" and raise warning
-
-        for df in self.get_DataFrame:
-            pass
-
+        if not ignore_columns:
+            ignore_columns = []
+
+        if unit_column_orig:
+            ignore_columns.append(unit_column_orig)
+            for entry in df_map.iterrows():
+                # need fullmatch here as the same is used in ioutil.convert
+                corresponding_rows = self.fullmatch(**entry[1].to_dict())
+                for row in corresponding_rows:
+                    if self.unit.loc[row].unit != entry[1][unit_column_orig]:
+                        raise ValueError(
+                            f"Unit in extension does not match the unit in mapping for row {row}"
+                            )
+        if unit_column_new:
+            ignore_columns.append(unit_column_new)
+
+
+        # for df in self.get_DataFrame:
+        # CONT:
+        # 
+        # 1) run with one and check output
+        # 2) check unit aggregation if unit_column_new is not given 
+        #   - unit aggregate with str join of unique entries
+        # 3) make dict and build new extension
+        #     # run convert of all dataframe and build new extension
+        #
 
 class IOSystem(BaseSystem):
     """Class containing a whole EE MRIO System

pymrio/tools/ioutil.py

@@ -1007,7 +1007,11 @@ def check_df_map(df_orig, df_map):
     pass
 
 
-def convert(df_orig, df_map, agg_func="sum", drop_not_bridged_index=True):
+def convert(df_orig, 
+            df_map, 
+            agg_func="sum", 
+            drop_not_bridged_index=True,
+            ignore_columns = None):
     """Convert a DataFrame to a new classification
 
     Parameters
@@ -1082,6 +1086,10 @@ def convert(df_orig, df_map, agg_func="sum", drop_not_bridged_index=True):
         make a bridge column for the ones to be dropped and map all to the same name.
         Then drop this index level after the conversion.
 
+    ignore_columns : list, optional
+        List of column names in df_map which should be ignored.
+        These could be columns with additional information, unit columns, etc.
+
 
     Extension for extensions:
     extension ... extension name
@@ -1098,6 +1106,9 @@ def convert(df_orig, df_map, agg_func="sum", drop_not_bridged_index=True):
     bridge_components = namedtuple("bridge_components", ["new", "orig", "raw"])
     bridges = []
 
+    if not ignore_columns:
+        ignore_columns = []
+
     if isinstance(df_orig, pd.Series):
         df_orig = pd.DataFrame(df_orig)
 
@@ -1125,7 +1136,8 @@ def convert(df_orig, df_map, agg_func="sum", drop_not_bridged_index=True):
     stacked_columns = []
     orig_column_index = df_orig.columns
     for col in df_map.columns:
-        if col in ["factor", "unit"]:
+        # TODO: other names for unit should be allowed
+        if col in (["factor"] + ignore_columns):
             continue
         if col not in df_orig.index.names:
             if col in df_orig.columns.names:

tests/test_core.py

@@ -552,6 +552,41 @@ def test_characterize_extension(fix_testmrio):
         check_names=False,
     )
 
+def test_extension_convert(fix_testmrio):
+
+    df_map = pd.DataFrame(
+        columns=["stressor", "compartment", "total__stressor", "factor", "unit_orig", "unit_new"],
+        data=[
+            ["emis.*", "air|water", "total_regex", 1000, "kg", "t"],
+            ["emission_type[1|2]", ".*", "total_sum", 1E-3, "kg", "g"],
+            ["emission_type1", ".*", "air_emissions", 1000, "kg", "t"],
+        ],
+    )
+
+    pass
+
+def test_extension_convert_test_unit(fix_testmrio):
+
+    df_fail1 = pd.DataFrame(
+        columns=["stressor", "compartment", "total__stressor", "factor", "unit_orig", "unit_new"],
+        data=[
+            ["emis.*", "air|water", "total_regex", 1000, "g", "t"],
+        ],
+    )
+
+    df_fail2 = pd.DataFrame(
+        columns=["stressor", "compartment", "total__stressor", "factor", "unit_emis", "unit_new"],
+        data=[
+            ["emission_type1", "air", "total_regex", 1000, "t", "t"],
+        ],
+    )
+
+    with pytest.raises(ValueError):
+        fix_testmrio.testmrio.emissions.convert(df_fail1, extension_name="emissions_new")
+
+    with pytest.raises(ValueError):
+        fix_testmrio.testmrio.emissions.convert(df_fail2, extension_name="emissions_new", unit_column_orig="unit_emis")
+
 
 def test_reset_to_flows(fix_testmrio):
     tt = fix_testmrio.testmrio