Added unit testing

.idea/Model.iml

@@ -2,7 +2,7 @@
 <module type="PYTHON_MODULE" version="4">
   <component name="NewModuleRootManager">
     <content url="file://$MODULE_DIR$" />
-    <orderEntry type="jdk" jdkName="Python 3.11 (hestia-data)" jdkType="Python SDK" />
+    <orderEntry type="jdk" jdkName="Python 3.10 (hestia-data)" jdkType="Python SDK" />
     <orderEntry type="sourceFolder" forTests="false" />
   </component>
 </module>

.idea/codeStyles/codeStyleConfig.xml

@@ -0,0 +1,5 @@
+<component name="ProjectCodeStyleConfiguration">
+  <state>
+    <option name="PREFERRED_PROJECT_CODE_STYLE" value="Default" />
+  </state>
+</component>

.idea/misc.xml

@@ -1,4 +1,7 @@
 <?xml version="1.0" encoding="UTF-8"?>
 <project version="4">
-  <component name="ProjectRootManager" version="2" project-jdk-name="Python 3.11 (hestia-data)" project-jdk-type="Python SDK" />
+  <component name="ProjectRootManager" version="2" project-jdk-name="Python 3.10 (hestia-data)" project-jdk-type="Python SDK" />
+  <component name="PythonCompatibilityInspectionAdvertiser">
+    <option name="version" value="3" />
+  </component>
 </project>

epc_data/EpcClean.py

@@ -2,6 +2,8 @@ import re
 from typing import List, Dict, Any, Union, Tuple
 from collections import Counter
 
+import pandas as pd
+
 
 class EpcClean:
     """
@@ -39,8 +41,12 @@ class EpcClean:
 
         for description in self.unique_vals["roof-description"].keys():
             self.cleaned["roof-description"].append(
-                {"original": description, "cleaned": self.clean_roof(description)}
+                {
+                    "original_description": description,
+                    **self.clean_roof(description)
+                }
             )
+        df = pd.DataFrame(self.cleaned["roof-description"])
 
     def _init_empty_cleaned_obj(self) -> None:
         """
@@ -181,8 +187,10 @@ class EpcClean:
             insulation_thickness = self._find_insulation_thickness(
                 description_lower, is_pitched, is_roof_room, is_flat
             )
+        elif description_lower == "pitched":
+            thermal_transmittence, thermal_transmittence_unit, insulation_thickness = None, None, None
         else:
-            raise NotImplementedError("Not handles this")
+            raise NotImplementedError("Not handled this")
 
         attributes = {
             "is_pitched": is_pitched,

epc_data/README.md

@@ -33,3 +33,17 @@ To install project dependencies navigate to /epc_data and run
 ```commandline
 pip install -r requirements.txt
 ```
+
+### Running Tests
+
+If you are not in a virtual environment, activate it with
+
+```commandline
+conda activate envName
+```
+
+Then run
+
+```commandline
+python -m pytest
+```

epc_data/app.py

@@ -1,3 +1,5 @@
+import pickle
+
 from tqdm import tqdm
 
 from epc_data.temp_inputs import input_data

epc_data/requirements.txt

@@ -2,4 +2,5 @@ epc-api-python
 python-dotenv
 tqdm
 pandas
-mypy
+mypy
+pytest

epc_data/tests/test_EpcClean.py

@@ -0,0 +1,132 @@
+import pytest
+import pickle
+from epc_data.EpcClean import EpcClean
+from pathlib import Path
+
+# For local testing
+if __file__ == "<input>":
+    input_data_path = Path("./epc_data/tests/test_data/EpcClean_inputs.obj")
+else:
+    current_file_path = Path(__file__)
+    input_data_path = current_file_path.parent / 'test_data' / 'EpcClean_inputs.obj'
+
+
+#
+# @pytest.fixture
+# def data():
+#     print("WOW")
+#     print(input_data_path)
+#     with open(input_data_path, 'rb') as f:
+#         data = pickle.load(f)
+#     return data
+#
+#
+# def test_clean(data):
+#     epc = EpcClean(data)
+#     epc.clean()
+#     assert len(epc.cleaned["roof-description"]) == len(epc.unique_vals["roof-description"])
+#
+#
+# def test_clean_roof(data):
+#     epc = EpcClean(data)
+#     result = epc.clean_roof('Pitched, 270 mm loft insulation')
+#
+#     # change the expected output based on your requirement
+#     expected_output = {
+#         "is_pitched": True,
+#         "is_roof_room": False,
+#         "has_loft": True,
+#         "insulation_thickness": 270,
+#         "has_dwelling_above": False,
+#         "assumed": False,
+#         "is_flat": False,
+#         "thermal_transmittence": None,
+#         "thermal_transmittence_unit": None
+#     }
+#
+#     assert result == expected_output
+#
+#
+# def test_clean_roof_with_dwelling_above(data):
+#     epc = EpcClean(data)
+#     result = epc.clean_roof('(another dwelling above)')
+#
+#     expected_output = {
+#         "is_pitched": False,
+#         "is_roof_room": False,
+#         "has_loft": False,
+#         "insulation_thickness": 0,
+#         "has_dwelling_above": True,
+#         "assumed": False,
+#         "is_flat": False,
+#         "is_thatched": False,
+#         "thermal_transmittence": None,
+#         "thermal_transmittence_unit": None,
+#     }
+#
+#     assert result == expected_output
+
+
+class TestEpcClean:
+    @pytest.fixture(autouse=True)
+    def load_data(self):
+        with open(input_data_path, "rb") as file:
+            self.data = pickle.load(file)
+        self.ec = EpcClean(self.data)
+
+    def test_clean(self):
+        self.ec.clean()
+        assert len(self.ec.cleaned["roof-description"]) == len(self.ec.unique_vals["roof-description"])
+
+    def test__init_empty_cleaned_obj(self):
+        self.ec._init_empty_cleaned_obj()
+        assert all([len(values) == 0 for values in self.ec.cleaned.values()])
+
+    def test__search_split_roof_description(self):
+        assert self.ec._search_split_roof_description("insulated") == "average"
+        assert self.ec._search_split_roof_description("limited") == "below average"
+        with pytest.raises(NotImplementedError):
+            self.ec._search_split_roof_description("unknown")
+
+    def test__find_insulation_thickness(self):
+        assert self.ec._find_insulation_thickness("no insulation", False, False, False) == 0
+
+    def test__extract_thermal_transmittence(self):
+        description = "U-value of 2.3 w/m-¦k"
+        assert self.ec._extract_thermal_transmittence(description) == (2.3, "w/m-¦k")
+
+    def test_clean_roof(self):
+        result = self.ec.clean_roof('Pitched, 270 mm loft insulation')
+
+        # change the expected output based on your requirement
+        expected_output = {
+            "is_pitched": True,
+            "is_roof_room": False,
+            "has_loft": True,
+            "insulation_thickness": 270,
+            "has_dwelling_above": False,
+            "assumed": False,
+            "is_flat": False,
+            "thermal_transmittence": None,
+            "thermal_transmittence_unit": None
+        }
+
+        assert result == expected_output
+
+    def test_clean_roof_with_dwelling_above(self):
+        result = self.ec.clean_roof('(another dwelling above)')
+
+        expected_output = {
+            "is_pitched": False,
+            "is_roof_room": False,
+            "has_loft": False,
+            "insulation_thickness": 0,
+            "has_dwelling_above": True,
+            "assumed": False,
+            "is_flat": False,
+            "is_thatched": False,
+            "thermal_transmittence": None,
+            "thermal_transmittence_unit": None,
+        }
+
+        assert result == expected_output