further epc clean debuggin

model_data/app.py

@@ -108,9 +108,9 @@ def app():
         # TODO: cleaner.cleaned datasets to s3
 
         # TODO: Add property age band into this
-        uvalue_estimates = UvalueEstimations(data=data)
-        uvalue_estimates.get_estimates(cleaner=cleaner)
-        # TODO: Store these to a s3
-        uvalue_estimates.walls
-        uvalue_estimates.floors
-        uvalue_estimates.roofs
+        # uvalue_estimates = UvalueEstimations(data=data)
+        # uvalue_estimates.get_estimates(cleaner=cleaner)
+        # # TODO: Store these to a s3
+        # uvalue_estimates.walls
+        # uvalue_estimates.floors
+        # uvalue_estimates.roofs

model_data/epc_attributes/FloorAttributes.py

@@ -1,3 +1,4 @@
+import re
 from typing import Dict, Union
 from model_data.BaseUtility import Definitions
 from model_data.epc_attributes.attribute_utils import extract_thermal_transmittance, extract_component_types
@@ -26,11 +27,7 @@ class FloorAttributes(Definitions):
             description in self.OBSERVED_ERRORS)
 
         # Try and perform a translation, incase it's in welsh
-        translation = self.WELSH_TEXT.get(self.description)
-
-        if translation:
-            self.nodata = False
-            self.description = translation
+        self.translate_welsh_text()
 
         if not self.nodata and not any(
             rt in self.description for rt in
@@ -38,6 +35,23 @@ class FloorAttributes(Definitions):
         ):
             raise ValueError('Invalid description')
 
+    def translate_welsh_text(self):
+
+        uvalue_match = re.search(
+            r'trawsyriannedd thermol cyfartalog (\d+(\.\d+)?)\s*w/m-¦k', self.description
+        )
+
+        # Step 2: Generalized translation with placeholder
+        if uvalue_match:
+            uvalue = uvalue_match.group(1)
+            self.description = f"average thermal transmittance {uvalue} W/m-¦K"
+
+        else:
+            translation = self.WELSH_TEXT.get(self.description)
+            if translation:
+                self.nodata = False
+                self.description = translation
+
     def process(self) -> Dict[str, Union[str, bool, int, None]]:
 
         if self.nodata:

model_data/epc_attributes/HotWaterAttributes.py

@@ -98,7 +98,8 @@ class HotWaterAttributes(Definitions):
         "ogçör brif system, adfer gwres nwyon ffliw": "from main system, flue gas heat recovery",
         "bwyler/cylchredydd nwy": "gas boiler/circulator",
         "ogçör brif system, dim thermostat ar y silindr": "from main system, no cylinder thermostat",
-        "twymwr tanddwr, an-frig": "electric immersion, off-peak"
+        "twymwr tanddwr, an-frig": "electric immersion, off-peak",
+        "ogçör brif system, gydag ynnigçör haul": "from main system, plus solar"
     }
 
     def __init__(self, description: str):

model_data/epc_attributes/MainheatAttributes.py

@@ -14,6 +14,7 @@ class MainHeatAttributes(Definitions):
         # buildings. The main output of micro-CHP systems is heat, with electricity generation as a secondary output.
         "micro-cogeneration",
         "solar assisted heat pump",
+        "exhaust source heat pump",
     ]
     FUEL_TYPES = ["electric", "mains gas", "wood logs", "LPG", "coal", "oil", "wood pellets", "anthracite",
                   "dual fuel mineral and wood", "smokeless fuel", "lpg"]

model_data/epc_attributes/RoofAttributes.py

@@ -11,7 +11,8 @@ class RoofAttributes(Definitions):
     WELSH_TEXT = {
         "ar oleddf, dim inswleiddio": "pitched, no insulation",
         "ar oleddf, dim inswleiddio (rhagdybiaeth)": "pitched, no insulation (assumed)",
-        "(annedd arall uwchben)": "(another dwelling above)"
+        "(annedd arall uwchben)": "(another dwelling above)",
+        "yn wastad, inswleiddio cyfyngedig (rhagdybiaeth)": "Flat, limited insulation (assumed)",
     }
 
     def __init__(self, description: str):

model_data/tests/test_data/test_hot_water_attributes_cases.py

@@ -145,5 +145,10 @@ hotwater_cases = [
     {'original_description': 'Twymwr tanddwr, an-frig', 'heater_type': 'electric immersion', 'system_type': None,
      'thermostat_characteristics': None, 'heating_scope': None, 'energy_recovery': None, 'tariff_type': 'off-peak',
      'extra_features': None, 'chp_systems': None, 'distribution_system': None, 'no_system_present': None,
-     'assumed': False, "appliance": None}
+     'assumed': False, "appliance": None},
+    {'original_description': 'OGÇÖr brif system, gydag ynniGÇÖr haul', 'heater_type': None,
+     'system_type': 'from main system',
+     'thermostat_characteristics': None, 'heating_scope': None, 'energy_recovery': None, 'tariff_type': None,
+     'extra_features': 'plus solar', 'chp_systems': None, 'distribution_system': None, 'no_system_present': None,
+     'assumed': False, "appliance": None},
 ]

model_data/tests/test_data/test_roof_attributes_cases.py

@@ -353,5 +353,9 @@ clean_roof_test_cases = [
     {'original_description': 'Ar oleddf, dim inswleiddio (rhagdybiaeth)', 'thermal_transmittance': None,
      'thermal_transmittance_unit': None, 'is_pitched': True, 'is_roof_room': False, 'is_loft': False, 'is_flat': False,
      'is_thatched': False, 'is_at_rafters': False, 'is_assumed': True, 'has_dwelling_above': False, 'is_valid': True,
-     'insulation_thickness': 'none'}
+     'insulation_thickness': 'none'},
+    {'original_description': 'Yn wastad, inswleiddio cyfyngedig (rhagdybiaeth)', 'thermal_transmittance': None,
+     'thermal_transmittance_unit': None, 'is_pitched': False, 'is_roof_room': False, 'is_loft': False, 'is_flat': True,
+     'is_thatched': False, 'is_at_rafters': False, 'is_assumed': True, 'has_dwelling_above': False, 'is_valid': True,
+     'insulation_thickness': 'below average'}
 ]