From 63c431796d730cb9ee4d3a4b47ac7ca226493569 Mon Sep 17 00:00:00 2001
From: Lukas Riedel <34276446+peanutfun@users.noreply.github.com>
Date: Tue, 4 Jul 2023 13:36:54 +0200
Subject: [PATCH 1/8] Remove all uses of Tag from climada/engine

---
 climada/engine/impact.py           | 114 ++++++-----------------------
 climada/engine/impact_data.py      |   9 ---
 climada/engine/test/test_impact.py |  40 +---------
 3 files changed, 22 insertions(+), 141 deletions(-)

diff --git a/climada/engine/impact.py b/climada/engine/impact.py
index 84840c613..39c97767d 100644
--- a/climada/engine/impact.py
+++ b/climada/engine/impact.py
@@ -51,7 +51,6 @@
 import climada.util.dates_times as u_dt
 import climada.util.plot as u_plot
 from climada.util.select import get_attributes_with_matching_dimension
-from climada.util.tag import Tag
 
 LOGGER = logging.getLogger(__name__)
 
@@ -88,9 +87,6 @@ class Impact():
     imp_mat : sparse.csr_matrix
         matrix num_events x num_exp with impacts.
         only filled if save_mat is True in calc()
-    tag : dict
-        dictionary of tags of exposures, impact functions set and
-        hazard: {'exp': Tag(), 'impf_set': Tag(), 'haz': Tag()}
     haz_type : str
         the hazard type of the hazard
     """
@@ -109,7 +105,6 @@ def __init__(self,
                  aai_agg=0,
                  unit='',
                  imp_mat=None,
-                 tag=None,
                  haz_type=''):
         """
         Init Impact object
@@ -145,15 +140,11 @@ def __init__(self,
             value unit used (given by exposures unit)
         imp_mat : sparse.csr_matrix, optional
             matrix num_events x num_exp with impacts.
-        tag : dict, optional
-            dictionary of tags of exposures, impact functions set and
-            hazard: {'exp': Tag(), 'impf_set': Tag(), 'haz': Tag()}
         haz_type : str, optional
             the hazard type
         """
 
         self.haz_type = haz_type
-        self.tag = tag or {}
         self.event_id = np.array([], int) if event_id is None else event_id
         self.event_name = [] if event_name is None else event_name
         self.date = np.array([], int) if date is None else date
@@ -260,10 +251,6 @@ def from_eih(cls, exposures, impfset, hazard,
             at_event = at_event,
             aai_agg = aai_agg,
             imp_mat = imp_mat if imp_mat is not None else sparse.csr_matrix((0, 0)),
-            tag = {'exp': exposures.tag,
-                   'impf_set': impfset.tag,
-                   'haz': hazard.tag
-                   },
             haz_type = hazard.haz_type,
             )
 
@@ -543,7 +530,6 @@ def calc_freq_curve(self, return_per=None):
             ifc_impact = interp_imp
 
         return ImpactFreqCurve(
-            tag=self.tag,
             return_per=ifc_return_per,
             impact=ifc_impact,
             unit=self.unit,
@@ -885,14 +871,10 @@ def write_csv(self, file_name):
         LOGGER.info('Writing %s', file_name)
         with open(file_name, "w", encoding='utf-8') as imp_file:
             imp_wr = csv.writer(imp_file)
-            imp_wr.writerow(["tag_hazard", "tag_exposure", "tag_impact_func",
-                             "unit", "tot_value", "aai_agg", "event_id",
+            imp_wr.writerow(["haz_type", "unit", "tot_value", "aai_agg", "event_id",
                              "event_name", "event_date", "event_frequency", "frequency_unit",
                              "at_event", "eai_exp", "exp_lat", "exp_lon", "exp_crs"])
-            csv_data = [[[self.haz_type], [self.tag['haz'].file_name],
-                         [self.tag['haz'].description]],
-                        [[self.tag['exp'].file_name], [self.tag['exp'].description]],
-                        [[self.tag['impf_set'].file_name], [self.tag['impf_set'].description]],
+            csv_data = [[self.haz_type],
                         [self.unit], [self._tot_value], [self.aai_agg],
                         self.event_id, self.event_name, self.date,
                         self.frequency, [self.frequency_unit], self.at_event,
@@ -920,32 +902,26 @@ def write_col(i_col, imp_ws, xls_data):
         imp_wb = xlsxwriter.Workbook(file_name)
         imp_ws = imp_wb.add_worksheet()
 
-        header = ["tag_hazard", "tag_exposure", "tag_impact_func",
-                  "unit", "tot_value", "aai_agg", "event_id",
+        header = ["haz_type", "unit", "tot_value", "aai_agg", "event_id",
                   "event_name", "event_date", "event_frequency", "frequency_unit",
                   "at_event", "eai_exp", "exp_lat", "exp_lon", "exp_crs"]
         for icol, head_dat in enumerate(header):
             imp_ws.write(0, icol, head_dat)
-        data = [str(self.haz_type), str(self.tag['haz'].file_name),
-                str(self.tag['haz'].description)]
+        data = [str(self.haz_type)]
         write_col(0, imp_ws, data)
-        data = [str(self.tag['exp'].file_name), str(self.tag['exp'].description)]
-        write_col(1, imp_ws, data)
-        data = [str(self.tag['impf_set'].file_name), str(self.tag['impf_set'].description)]
-        write_col(2, imp_ws, data)
-        write_col(3, imp_ws, [self.unit])
-        write_col(4, imp_ws, [self._tot_value])
-        write_col(5, imp_ws, [self.aai_agg])
-        write_col(6, imp_ws, self.event_id)
-        write_col(7, imp_ws, self.event_name)
-        write_col(8, imp_ws, self.date)
-        write_col(9, imp_ws, self.frequency)
-        write_col(10, imp_ws, [self.frequency_unit])
-        write_col(11, imp_ws, self.at_event)
-        write_col(12, imp_ws, self.eai_exp)
-        write_col(13, imp_ws, self.coord_exp[:, 0])
-        write_col(14, imp_ws, self.coord_exp[:, 1])
-        write_col(15, imp_ws, [str(self.crs)])
+        write_col(1, imp_ws, [self.unit])
+        write_col(2, imp_ws, [self._tot_value])
+        write_col(3, imp_ws, [self.aai_agg])
+        write_col(4, imp_ws, self.event_id)
+        write_col(5, imp_ws, self.event_name)
+        write_col(6, imp_ws, self.date)
+        write_col(7, imp_ws, self.frequency)
+        write_col(8, imp_ws, [self.frequency_unit])
+        write_col(9, imp_ws, self.at_event)
+        write_col(10, imp_ws, self.eai_exp)
+        write_col(11, imp_ws, self.coord_exp[:, 0])
+        write_col(12, imp_ws, self.coord_exp[:, 1])
+        write_col(13, imp_ws, [str(self.crs)])
 
         imp_wb.close()
 
@@ -1023,11 +999,6 @@ def write_dict(group, name, value):
             for key, val in value.items():
                 write(group, key, val)
 
-        def write_tag(group, name, value):
-            """Write a tag object using the dict writer"""
-            group = group.create_group(name)  # name is 'exp', 'haz', 'impf_set'
-            value.to_hdf5(group)              # value is a Tag
-
         def _write_csr_dense(group, name, value):
             """Write a CSR Matrix in dense format"""
             group.create_dataset(name, data=value.toarray())
@@ -1052,7 +1023,6 @@ def write_csr(group, name, value):
         #       2) Anything is 'object', so this serves as fallback/default.
         type_writers = {
             str: write_attribute,
-            Tag: write_tag,
             dict: write_dict,
             sparse.csr_matrix: write_csr,
             Collection: write_dataset,
@@ -1107,7 +1077,7 @@ def from_csv(cls, file_name):
         # pylint: disable=no-member
         LOGGER.info('Reading %s', file_name)
         imp_df = pd.read_csv(file_name)
-        imp = cls(haz_type=str(imp_df.tag_hazard[0]))
+        imp = cls(haz_type=imp_df.haz_type[0])
         imp.unit = imp_df.unit[0]
         imp.tot_value = imp_df.tot_value[0]
         imp.aai_agg = imp_df.aai_agg[0]
@@ -1128,12 +1098,7 @@ def from_csv(cls, file_name):
             imp.crs = u_coord.to_crs_user_input(imp_df.exp_crs.values[0])
         except AttributeError:
             imp.crs = DEF_CRS
-        imp.tag['haz'] = Tag(str(imp_df.tag_hazard[1]),
-                             str(imp_df.tag_hazard[2]))
-        imp.tag['exp'] = Tag(str(imp_df.tag_exposure[0]),
-                             str(imp_df.tag_exposure[1]))
-        imp.tag['impf_set'] = Tag(str(imp_df.tag_impact_func[0]),
-                                  str(imp_df.tag_impact_func[1]))
+
         return imp
 
     def read_csv(self, *args, **kwargs):
@@ -1158,16 +1123,7 @@ def from_excel(cls, file_name):
         """
         LOGGER.info('Reading %s', file_name)
         dfr = pd.read_excel(file_name)
-        imp = cls(haz_type=str(dfr['tag_hazard'][0]))
-        imp.tag['haz'] = Tag(
-            file_name = dfr['tag_hazard'][1],
-            description = dfr['tag_hazard'][2])
-        imp.tag['exp'] = Tag()
-        imp.tag['exp'].file_name = dfr['tag_exposure'][0]
-        imp.tag['exp'].description = dfr['tag_exposure'][1]
-        imp.tag['impf_set'] = Tag()
-        imp.tag['impf_set'].file_name = dfr['tag_impact_func'][0]
-        imp.tag['impf_set'].description = dfr['tag_impact_func'][1]
+        imp = cls(haz_type=str(dfr['haz_type'][0]))
 
         imp.unit = dfr.unit[0]
         imp.tot_value = dfr.tot_value[0]
@@ -1215,20 +1171,6 @@ def from_hdf5(cls, file_path: Union[str, Path]):
             ├─ event_name
             ├─ frequency
             ├─ imp_mat
-            ├─ tag/
-            │  ├─ exp/
-            │  │  ├─ .attrs/
-            │  │  │  ├─ file_name
-            │  │  │  ├─ description
-            │  ├─ haz/
-            │  │  ├─ .attrs/
-            │  │  │  ├─ haz_type
-            │  │  │  ├─ file_name
-            │  │  │  ├─ description
-            │  ├─ impf_set/
-            │  │  ├─ .attrs/
-            │  │  │  ├─ file_name
-            │  │  │  ├─ description
             ├─ .attrs/
             │  ├─ aai_agg
             │  ├─ crs
@@ -1300,13 +1242,6 @@ def from_hdf5(cls, file_path: Union[str, Path]):
                 # pylint: disable=no-member
                 kwargs["event_name"] = list(file["event_name"].asstr()[:])
 
-            # Tags
-            if "tag" in file:
-                tag_group = file["tag"]
-                # the tag group has tags for 'exp', 'haz' and 'impf_set'
-                tag_kwargs = {tag: Tag.from_hdf5(tag_group[tag]) for tag in tag_group.keys()}
-
-                kwargs["tag"] = tag_kwargs
         # Create the impact object
         return cls(**kwargs)
 
@@ -1476,7 +1411,6 @@ def _build_exp(self):
             crs=self.crs,
             value_unit=self.unit,
             ref_year=0,
-            tag=Tag(),
             meta=None
         )
 
@@ -1498,7 +1432,6 @@ def _build_exp_event(self, event_id):
             crs=self.crs,
             value_unit=self.unit,
             ref_year=0,
-            tag=Tag(),
             meta=None
         )
 
@@ -1719,7 +1652,7 @@ def concat(cls, imp_list: Iterable, reset_event_ids: bool = False):
           ``frequency``, ``imp_mat``, ``at_event``,
         - sums up the values of attributes ``eai_exp``, ``aai_exp``
         - and takes the following attributes from the first impact object in the passed
-          impact list: ``coord_exp``, ``crs``, ``unit``, ``tot_value``, ``tag``,
+          impact list: ``coord_exp``, ``crs``, ``unit``, ``tot_value``,
           ``frequency_unit``
 
         If event ids are not unique among the passed impact objects an error is raised.
@@ -1806,7 +1739,6 @@ def stack_attribute(attr_name: str) -> np.ndarray:
             eai_exp=np.nansum([imp.eai_exp for imp in imp_list], axis=0),
             aai_agg=np.nansum([imp.aai_agg for imp in imp_list]),
             imp_mat=imp_mat,
-            tag=first_imp.tag,
             haz_type=first_imp.haz_type,
             frequency_unit=first_imp.frequency_unit,
             **kwargs,
@@ -1850,10 +1782,6 @@ class ImpactFreqCurve():
     """Impact exceedence frequency curve.
     """
 
-    tag : dict = field(default_factory=dict)
-    """dictionary of tags of exposures, impact functions set and
-        hazard: {'exp': Tag(), 'impf_set': Tag(), 'haz': Tag()}"""
-
     return_per : np.array = np.array([])
     """return period"""
 
diff --git a/climada/engine/impact_data.py b/climada/engine/impact_data.py
index 2b5bc8410..f8cde2838 100644
--- a/climada/engine/impact_data.py
+++ b/climada/engine/impact_data.py
@@ -30,7 +30,6 @@
 from climada.util.constants import DEF_CRS
 import climada.util.coordinates as u_coord
 from climada.engine import Impact
-from climada.util.tag import Tag
 
 LOGGER = logging.getLogger(__name__)
 
@@ -956,14 +955,6 @@ def emdat_to_impact(emdat_file_csv, hazard_type_climada, year_range=None, countr
     # Inititate Impact-instance:
     impact_instance = Impact(haz_type=hazard_type_climada)
 
-    impact_instance.tag = dict()
-    impact_instance.tag['haz'] = Tag(file_name=emdat_file_csv,
-                                     description='EM-DAT impact, direct import')
-    impact_instance.tag['exp'] = Tag(file_name=emdat_file_csv,
-                                     description='EM-DAT impact, direct import')
-    impact_instance.tag['impf_set'] = Tag(file_name=None, description=None)
-
-
     # Load EM-DAT impact data by event:
     em_data = emdat_impact_event(emdat_file_csv, countries=countries, hazard=hazard_type_emdat,
                                  year_range=year_range, reference_year=reference_year,
diff --git a/climada/engine/test/test_impact.py b/climada/engine/test/test_impact.py
index 70b870fe2..4db50d6c4 100644
--- a/climada/engine/test/test_impact.py
+++ b/climada/engine/test/test_impact.py
@@ -28,7 +28,6 @@
 from pyproj import CRS
 from rasterio.crs import CRS as rCRS
 
-from climada.util.tag import Tag
 from climada.entity.entity_def import Entity
 from climada.hazard.base import Hazard
 from climada.engine import Impact, ImpactCalc
@@ -65,11 +64,6 @@ def dummy_impact():
         imp_mat=sparse.csr_matrix(
             np.array([[0, 0], [1, 1], [2, 2], [3, 3], [30, 30], [31, 31]])
         ),
-        tag={
-            "exp": Tag("file_exp.p", "descr exp"),
-            "haz": Tag("file_haz.p", "descr haz"),
-            "impf_set": Tag(),
-        },
         haz_type="TC",
     )
 
@@ -370,9 +364,6 @@ def test_write_read_ev_test(self):
         num_ev = 10
         num_exp = 5
         imp_write = Impact(haz_type='TC')
-        imp_write.tag = {'exp': Tag('file_exp.p', 'descr exp'),
-                         'haz': Tag('file_haz.p', 'descr haz'),
-                         'impf_set': Tag()}
         imp_write.event_id = np.arange(num_ev)
         imp_write.event_name = ['event_' + str(num) for num in imp_write.event_id]
         imp_write.date = np.ones(num_ev)
@@ -410,9 +401,6 @@ def test_write_read_exp_test(self):
         num_ev = 5
         num_exp = 10
         imp_write = Impact(haz_type='TC')
-        imp_write.tag = {'exp': Tag('file_exp.p', 'descr exp'),
-                         'haz': Tag('file_haz.p', 'descr haz'),
-                         'impf_set': Tag()}
         imp_write.event_id = np.arange(num_ev)
         imp_write.event_name = ['event_' + str(num) for num in imp_write.event_id]
         imp_write.date = np.ones(num_ev)
@@ -950,11 +938,6 @@ def _compare_file_to_imp(self, filepath, impact, dense_imp_mat):
             self.assertEqual(file.attrs["aai_agg"], impact.aai_agg)
             self.assertEqual(file.attrs["frequency_unit"], impact.frequency_unit)
 
-            for tagtype in ["exp", "haz", "impf_set"]:
-                self.assertDictEqual(
-                    Tag.from_hdf5(file["tag"][tagtype]).__dict__, impact.tag[tagtype].__dict__
-                )
-
             if dense_imp_mat:
                 npt.assert_array_equal(file["imp_mat"], impact.imp_mat.toarray())
             else:
@@ -972,11 +955,7 @@ def _compare_impacts(self, impact_1, impact_2):
         for name, value in impact_1.__dict__.items():
             self.assertIn(name, impact_2.__dict__)
             value_comp = getattr(impact_2, name)
-            # NOTE: Tags do not compare
-            if name == "tag":
-                for key in value:
-                    self.assertDictEqual(value[key].__dict__, value_comp[key].__dict__)
-            elif isinstance(value, sparse.csr_matrix):
+            if isinstance(value, sparse.csr_matrix):
                 npt.assert_array_equal(value.toarray(), value_comp.toarray())
             elif np.ndim(value) > 0:
                 npt.assert_array_equal(value, value_comp)
@@ -1035,7 +1014,6 @@ def test_read_hdf5_minimal(self):
         self.assertEqual(impact.tot_value, 0)
         self.assertEqual(impact.aai_agg, 0)
         self.assertEqual(impact.unit, "")
-        self.assertEqual(impact.tag, {})
         self.assertEqual(impact.haz_type, "")
 
     def test_read_hdf5_full(self):
@@ -1055,14 +1033,6 @@ def test_read_hdf5_full(self):
         aai_agg = 200
         unit = "unit"
         haz_type="haz_type"
-        haz_tag = dict(file_name=["file_name"], description=["description"])
-        exp_tag = dict(file_name=["exp"], description=["exp"])
-        impf_set_tag = dict(file_name=["impf_set"], description=["impf_set"])
-
-        def write_tag(group, tag_kwds):
-            for key, value in tag_kwds.items():
-                array = group.create_dataset(key, (1,0), STR_DT)
-                array[0] = value
 
         # Write the data
         with h5py.File(self.filepath, "w") as file:
@@ -1081,11 +1051,6 @@ def write_tag(group, tag_kwds):
             file.attrs["tot_value"] = tot_value
             file.attrs["aai_agg"] = aai_agg
             file.attrs["unit"] = unit
-            for group, kwds in zip(
-                ("haz", "exp", "impf_set"), (haz_tag, exp_tag, impf_set_tag)
-            ):
-                taghdf5 = file.create_group(f"tag/{group}")
-                Tag(**kwds).to_hdf5(taghdf5)
             file.attrs["haz_type"] = haz_type
 
         # Load and check
@@ -1104,9 +1069,6 @@ def write_tag(group, tag_kwds):
         self.assertEqual(impact.tot_value, tot_value)
         self.assertEqual(impact.aai_agg, aai_agg)
         self.assertEqual(impact.unit, unit)
-        self.assertEqual(impact.tag["haz"].__dict__, haz_tag)
-        self.assertEqual(impact.tag["exp"].__dict__, exp_tag)
-        self.assertEqual(impact.tag["impf_set"].__dict__, impf_set_tag)
         self.assertEqual(impact.haz_type, haz_type)
 
         # Check with sparse

From 8f2c707e02f1e0f417ca0b7d6d967192fffb6b96 Mon Sep 17 00:00:00 2001
From: Lukas Riedel <34276446+peanutfun@users.noreply.github.com>
Date: Tue, 4 Jul 2023 13:47:56 +0200
Subject: [PATCH 2/8] Remove tag usage from yearsets and measures tests

---
 climada/entity/measures/test/test_base.py | 6 ------
 climada/util/yearsets.py                  | 2 --
 2 files changed, 8 deletions(-)

diff --git a/climada/entity/measures/test/test_base.py b/climada/entity/measures/test/test_base.py
index c20df313a..7b6d7b399 100644
--- a/climada/entity/measures/test/test_base.py
+++ b/climada/entity/measures/test/test_base.py
@@ -397,9 +397,6 @@ def test_calc_impact_pass(self):
         self.assertTrue(np.array_equal(imp.event_id, hazard.event_id))
         self.assertTrue(np.array_equal(imp.date, hazard.date))
         self.assertEqual(imp.event_name, hazard.event_name)
-        self.assertEqual(imp.tag['exp'].file_name, entity.exposures.tag.file_name)
-        self.assertEqual(imp.tag['haz'].file_name, hazard.tag.file_name)
-        self.assertEqual(imp.tag['impf_set'].file_name, entity.impact_funcs.tag.file_name)
         self.assertEqual(risk_transf.aai_agg, 0)
 
 
@@ -438,9 +435,6 @@ def test_calc_impact_transf_pass(self):
         self.assertTrue(np.array_equal(imp.event_id, hazard.event_id))
         self.assertTrue(np.array_equal(imp.date, hazard.date))
         self.assertEqual(imp.event_name, hazard.event_name)
-        self.assertEqual(imp.tag['exp'].file_name, entity.exposures.tag.file_name)
-        self.assertEqual(imp.tag['haz'].file_name, hazard.tag.file_name)
-        self.assertEqual(imp.tag['impf_set'].file_name, entity.impact_funcs.tag.file_name)
         self.assertEqual(risk_transf.aai_agg, 2.3139691495470852e+08)
 
 # Execute Tests
diff --git a/climada/util/yearsets.py b/climada/util/yearsets.py
index d32141dca..0c3b0033f 100755
--- a/climada/util/yearsets.py
+++ b/climada/util/yearsets.py
@@ -82,7 +82,6 @@ def impact_yearset(imp, sampled_years, lam=None, correction_fac=True, seed=None)
 
     #save calculations in yimp
     yimp.event_id = np.arange(1, n_sampled_years+1)
-    yimp.tag['yimp object'] = True
     yimp.date = u_dt.str_to_date([str(date) + '-01-01' for date in sampled_years])
     yimp.frequency = np.ones(n_sampled_years)*sum(len(row) for row in sampling_vect
                                                             )/n_sampled_years
@@ -139,7 +138,6 @@ def impact_yearset_from_sampling_vect(imp, sampled_years, sampling_vect, correct
     yimp.at_event = imp_per_year
     n_sampled_years = len(sampled_years)
     yimp.event_id = np.arange(1, n_sampled_years+1)
-    yimp.tag['yimp object'] = True
     yimp.date = u_dt.str_to_date([str(date) + '-01-01' for date in sampled_years])
     yimp.frequency = np.ones(n_sampled_years)*sum(len(row) for row in sampling_vect
                                                             )/n_sampled_years

From f8298102515f0c683c14ba4d193eb1fabdfafb05 Mon Sep 17 00:00:00 2001
From: Lukas Riedel <34276446+peanutfun@users.noreply.github.com>
Date: Tue, 4 Jul 2023 13:58:15 +0200
Subject: [PATCH 3/8] Update CHANGELOG.md

---
 CHANGELOG.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index d7ab9e784..2bb90c2c4 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -68,8 +68,8 @@ Removed:
 ### Removed
 
 - `Centroids.set_raster_from_pix_bounds` [#721](https://github.com/CLIMADA-project/climada_python/pull/721)
-
 - `requirements/env_developer.yml` environment specs. Use 'extra' requirements when installing the Python package instead [#712](https://github.com/CLIMADA-project/climada_python/pull/712)
+- `Impact.tag` attribute. This change is not backwards-compatible with respect to the files written and read by the `Impact` class [#743](https://github.com/CLIMADA-project/climada_python/pull/743)
 
 ## v3.3.2
 

From de5704e07a353d10063eb647d990f1f9d4d990ef Mon Sep 17 00:00:00 2001
From: Lukas Riedel <34276446+peanutfun@users.noreply.github.com>
Date: Wed, 5 Jul 2023 11:03:31 +0200
Subject: [PATCH 4/8] Make sure Impact.haz_type is cycled through H5

---
 climada/engine/impact.py           | 1 +
 climada/engine/test/test_impact.py | 1 +
 2 files changed, 2 insertions(+)

diff --git a/climada/engine/impact.py b/climada/engine/impact.py
index 39c97767d..845b658e6 100644
--- a/climada/engine/impact.py
+++ b/climada/engine/impact.py
@@ -1175,6 +1175,7 @@ def from_hdf5(cls, file_path: Union[str, Path]):
             │  ├─ aai_agg
             │  ├─ crs
             │  ├─ frequency_unit
+            │  ├─ haz_type
             │  ├─ tot_value
             │  ├─ unit
 
diff --git a/climada/engine/test/test_impact.py b/climada/engine/test/test_impact.py
index 4db50d6c4..0ab8739a1 100644
--- a/climada/engine/test/test_impact.py
+++ b/climada/engine/test/test_impact.py
@@ -937,6 +937,7 @@ def _compare_file_to_imp(self, filepath, impact, dense_imp_mat):
             self.assertEqual(file.attrs["unit"], impact.unit)
             self.assertEqual(file.attrs["aai_agg"], impact.aai_agg)
             self.assertEqual(file.attrs["frequency_unit"], impact.frequency_unit)
+            self.assertEqual(file.attrs["haz_type"], impact.haz_type)
 
             if dense_imp_mat:
                 npt.assert_array_equal(file["imp_mat"], impact.imp_mat.toarray())

From b896a4fceb5b813a6236ded3188d25a6d80778a4 Mon Sep 17 00:00:00 2001
From: Lukas Riedel <34276446+peanutfun@users.noreply.github.com>
Date: Wed, 5 Jul 2023 11:04:03 +0200
Subject: [PATCH 5/8] Remove unused import

---
 climada/engine/impact.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/climada/engine/impact.py b/climada/engine/impact.py
index 845b658e6..726e68897 100644
--- a/climada/engine/impact.py
+++ b/climada/engine/impact.py
@@ -21,7 +21,7 @@
 
 __all__ = ['ImpactFreqCurve', 'Impact']
 
-from dataclasses import dataclass, field
+from dataclasses import dataclass
 import logging
 import copy
 import csv

From 4996c0ec1cf0d62be75c22d2764cc3f3d0c71387 Mon Sep 17 00:00:00 2001
From: Lukas Riedel <34276446+peanutfun@users.noreply.github.com>
Date: Wed, 5 Jul 2023 11:05:39 +0200
Subject: [PATCH 6/8] Update climada/engine/impact.py

Co-authored-by: Chahan M. Kropf <chahan.kropf@usys.ethz.ch>
---
 climada/engine/impact.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/climada/engine/impact.py b/climada/engine/impact.py
index 39c97767d..0621e7fd2 100644
--- a/climada/engine/impact.py
+++ b/climada/engine/impact.py
@@ -1653,7 +1653,7 @@ def concat(cls, imp_list: Iterable, reset_event_ids: bool = False):
         - sums up the values of attributes ``eai_exp``, ``aai_exp``
         - and takes the following attributes from the first impact object in the passed
           impact list: ``coord_exp``, ``crs``, ``unit``, ``tot_value``,
-          ``frequency_unit``
+          ``frequency_unit``, ``haz_type``
 
         If event ids are not unique among the passed impact objects an error is raised.
         In this case, the user can set ``reset_event_ids=True`` to create unique event ids

From 3ca650cf0d005b25a2dc38941e5586e29970ab8a Mon Sep 17 00:00:00 2001
From: Lukas Riedel <34276446+peanutfun@users.noreply.github.com>
Date: Mon, 10 Jul 2023 09:48:27 +0200
Subject: [PATCH 7/8] Remove `impfset` argument from `Impact.from_eih`

---
 climada/engine/impact.py                |  6 ++++--
 climada/engine/impact_calc.py           | 10 +++++-----
 climada/engine/test/test_impact.py      |  6 ++----
 climada/engine/test/test_impact_calc.py | 10 ++++------
 4 files changed, 15 insertions(+), 17 deletions(-)

diff --git a/climada/engine/impact.py b/climada/engine/impact.py
index 42cf2bc7e..c8ec350f8 100644
--- a/climada/engine/impact.py
+++ b/climada/engine/impact.py
@@ -207,11 +207,13 @@ def calc(self, exposures, impact_funcs, hazard, save_mat=False, assign_centroids
 
 #TODO: new name
     @classmethod
-    def from_eih(cls, exposures, impfset, hazard,
-                 at_event, eai_exp, aai_agg, imp_mat=None):
+    def from_eih(cls, exposures, hazard, at_event, eai_exp, aai_agg, imp_mat=None):
         """
         Set Impact attributes from precalculated impact metrics.
 
+        .. versionchanged:: 3.3
+           The ``impfset`` argument was removed.
+
         Parameters
         ----------
         exposures : climada.entity.Exposures
diff --git a/climada/engine/impact_calc.py b/climada/engine/impact_calc.py
index 6040a5731..40bec773d 100644
--- a/climada/engine/impact_calc.py
+++ b/climada/engine/impact_calc.py
@@ -175,9 +175,8 @@ def _return_impact(self, imp_mat_gen, save_mat):
             imp_mat = None
             at_event, eai_exp, aai_agg = self.stitch_risk_metrics(imp_mat_gen)
         return Impact.from_eih(
-            self.exposures, self.impfset, self.hazard,
-            at_event, eai_exp, aai_agg, imp_mat
-            )
+            self.exposures, self.hazard, at_event, eai_exp, aai_agg, imp_mat
+        )
 
     def _return_empty(self, save_mat):
         """
@@ -202,8 +201,9 @@ def _return_empty(self, save_mat):
                 )
         else:
             imp_mat = None
-        return Impact.from_eih(self.exposures, self.impfset, self.hazard,
-                        at_event, eai_exp, aai_agg, imp_mat)
+        return Impact.from_eih(
+            self.exposures, self.hazard, at_event, eai_exp, aai_agg, imp_mat
+        )
 
     def minimal_exp_gdf(self, impf_col, assign_centroids, ignore_cover, ignore_deductible):
         """Get minimal exposures geodataframe for impact computation
diff --git a/climada/engine/test/test_impact.py b/climada/engine/test/test_impact.py
index 0ab8739a1..698a8bcd6 100644
--- a/climada/engine/test/test_impact.py
+++ b/climada/engine/test/test_impact.py
@@ -77,8 +77,7 @@ def test_from_eih_pass(self):
         fake_eai_exp = np.arange(len(exp.gdf))
         fake_at_event = np.arange(HAZ.size)
         fake_aai_agg = np.sum(fake_eai_exp)
-        imp = Impact.from_eih(exp, ENT.impact_funcs, HAZ,
-                              fake_at_event, fake_eai_exp, fake_aai_agg)
+        imp = Impact.from_eih(exp, HAZ, fake_at_event, fake_eai_exp, fake_aai_agg)
         self.assertEqual(imp.crs, exp.crs)
         self.assertEqual(imp.aai_agg, fake_aai_agg)
         self.assertEqual(imp.imp_mat.size, 0)
@@ -900,8 +899,7 @@ def test_match_centroids(self):
         fake_eai_exp = np.arange(len(exp.gdf))
         fake_at_event = np.arange(HAZ.size)
         fake_aai_agg = np.sum(fake_eai_exp)
-        imp = Impact.from_eih(exp, ENT.impact_funcs, HAZ,
-                              fake_at_event, fake_eai_exp, fake_aai_agg)
+        imp = Impact.from_eih(exp, HAZ, fake_at_event, fake_eai_exp, fake_aai_agg)
         imp_centr = imp.match_centroids(HAZ)
         np.testing.assert_array_equal(imp_centr, exp.gdf.centr_TC)
 
diff --git a/climada/engine/test/test_impact_calc.py b/climada/engine/test/test_impact_calc.py
index 65f2925d8..68b61c39a 100644
--- a/climada/engine/test/test_impact_calc.py
+++ b/climada/engine/test/test_impact_calc.py
@@ -686,7 +686,6 @@ def test_save_mat(self, from_eih_mock):
         self.icalc._return_impact(self.imp_mat_gen, save_mat=True)
         from_eih_mock.assert_called_once_with(
             ENT.exposures,
-            ENT.impact_funcs,
             HAZ,
             "at_event",
             "eai_exp",
@@ -707,11 +706,10 @@ def test_skip_mat(self, from_eih_mock):
         # Need to check every argument individually due to the last one being a matrix
         call_args = from_eih_mock.call_args.args
         self.assertEqual(call_args[0], ENT.exposures)
-        self.assertEqual(call_args[1], ENT.impact_funcs)
-        self.assertEqual(call_args[2], HAZ)
-        self.assertEqual(call_args[3], "at_event")
-        self.assertEqual(call_args[4], "eai_exp")
-        self.assertEqual(call_args[5], "aai_agg")
+        self.assertEqual(call_args[1], HAZ)
+        self.assertEqual(call_args[2], "at_event")
+        self.assertEqual(call_args[3], "eai_exp")
+        self.assertEqual(call_args[4], "aai_agg")
         np.testing.assert_array_equal(
             from_eih_mock.call_args.args[-1], sparse.csr_matrix((0, 0)).toarray()
         )

From 7b6f65d191cbdf5845b346ec25e35554d4f4385a Mon Sep 17 00:00:00 2001
From: Lukas Riedel <34276446+peanutfun@users.noreply.github.com>
Date: Mon, 10 Jul 2023 09:54:39 +0200
Subject: [PATCH 8/8] Remove mention of tag in impact tutorial

---
 doc/tutorial/climada_engine_Impact.ipynb | 1 -
 1 file changed, 1 deletion(-)

diff --git a/doc/tutorial/climada_engine_Impact.ipynb b/doc/tutorial/climada_engine_Impact.ipynb
index 1250ce1df..7ed751e41 100644
--- a/doc/tutorial/climada_engine_Impact.ipynb
+++ b/doc/tutorial/climada_engine_Impact.ipynb
@@ -65,7 +65,6 @@
    "source": [
     "| Attributes from input | Data Type | Description|\n",
     "| :- | :- | :- |\n",
-    "| tag |(dict)| dictionary storing the tags of the inputs (Exposure.tag, ImpactFuncSet.tag Hazard.tag)|\n",
     "| event_id |list(int)| id (>0) of each hazard event (Hazard.event_id)|\n",
     "| event_name |(list(str))| name of each event (Hazard.event_name)|\n",
     "| date |np.array| date of events (Hazard.date)|\n",