test - that data is clean

tests/unit/test_clean_data.py

@@ -0,0 +1,103 @@
+import unittest
+import pandas as pd
+import numpy as np
+from src.my_package.util import (
+    kelvin_to_celsius, 
+    ensure_rain_column, 
+    ensure_snow_column, 
+    fill_rain_column, 
+    fill_snow_column, 
+    extract_city_df
+)
+
+class TestUtilFunctions(unittest.TestCase):
+
+    # Test kelvin to celsius conversion
+    def test_kelvin_to_celsius(self):
+        kelvin_temp = 300
+        expected_celsius = 26.85
+
+        result = kelvin_to_celsius(kelvin_temp)
+
+        self.assertAlmostEqual(result, expected_celsius, places=2)
+
+    # Test if 'rain.1h' column is added when not present
+    def test_ensure_rain_column(self):
+        df = pd.DataFrame({
+            'temp': [300, 302, 305],
+            'humidity': [80, 82, 78]
+        })
+
+        df = ensure_rain_column(df)
+
+        # Check if the 'rain.1h' column is present after function call
+        self.assertTrue('rain.1h' in df.columns)
+        # Check if the column has NaN values
+        self.assertTrue(df['rain.1h'].isna().all())
+
+    # Test if 'snow.1h' column is added when not present
+    def test_ensure_snow_column(self):
+        df = pd.DataFrame({
+            'temp': [300, 302, 305],
+            'humidity': [80, 82, 78]
+        })
+
+        df = ensure_snow_column(df)
+
+        # Check if the 'snow.1h' column is present after function call
+        self.assertTrue('snow.1h' in df.columns)
+        # Check if the column has NaN values
+        self.assertTrue(df['snow.1h'].isna().all())
+
+    # Test if NaN values in 'rain.1h' are filled with 0
+    def test_fill_rain_column(self):
+        df = pd.DataFrame({
+            'temp': [300, 302, 305],
+            'rain.1h': [np.nan, 1.0, np.nan]
+        })
+
+        df = fill_rain_column(df)
+
+        # Check if NaN values are replaced with 0
+        self.assertEqual(df['rain.1h'].iloc[0], 0)
+        self.assertEqual(df['rain.1h'].iloc[2], 0)
+
+    # Test if NaN values in 'snow.1h' are filled with 0
+    def test_fill_snow_column(self):
+        df = pd.DataFrame({
+            'temp': [300, 302, 305],
+            'snow.1h': [np.nan, 0.5, np.nan]
+        })
+
+        df = fill_snow_column(df)
+
+        # Check if NaN values are replaced with 0
+        self.assertEqual(df['snow.1h'].iloc[0], 0)
+        self.assertEqual(df['snow.1h'].iloc[2], 0)
+
+    # Test extracting city DataFrame from JSON data
+    def test_extract_city_df(self):
+        weather_data = {
+            "list": [
+                {"dt": 1618245600, "temp": 290, "humidity": 85, "weather": [{"description": "clear sky"}]},
+                {"dt": 1618255600, "temp": 295, "humidity": 80, "weather": [{"description": "cloudy"}]},
+                {"dt": 1618265600, "temp": 300, "humidity": 75, "weather": [{"description": "sunny"}]}
+            ]
+        }
+
+        df = extract_city_df(weather_data)
+
+        # Ensure that 'dt' is the index and in datetime format
+        self.assertTrue(pd.api.types.is_datetime64_any_dtype(df.index))
+
+        # Check if duplicates are removed based on 'dt'
+        self.assertEqual(len(df), 3) 
+
+        # Check if 'weather' column was dropped
+        self.assertNotIn('weather', df.columns)
+
+        # Check the correct conversion of 'dt' to datetime
+        self.assertEqual(df.index[0], pd.to_datetime(1618245600, unit='s'))
+
+if __name__ == '__main__':
+    unittest.main()