Feature/test pandas spc x calc

tests/unittests/test_pandas_spc_x_calc.py

@@ -0,0 +1,116 @@
+# Python source
+# -------------------------------------------------------------------------
+# Copyright (c) 2023 NHS Python Community. All rights reserved.
+# Licensed under the MIT License. See license.txt in the project root for
+# license information.
+# -------------------------------------------------------------------------
+
+# FILE:           test_pandas_spc_x_calc.py
+
+# DESCRIPTION:    Tests on the pandas_scp_x_calc function.Given a pandas DataFrame,
+#                  a string indicating the column name of the values to be analysed,,
+#                 and an optional integer representing the number of values after which
+#                 the mean and other calculations should be fixed. It returns a pandas
+#                 DataFrame with the same values and the Statistic Process Control (SPC) values.
+#                 The SCP values const of:
+#                   - mean: The mean of the input values
+#                   - lpl: The lower process limit of the input values
+#                   - upl: The upper process limit of the input values
+#                   - outside_limits: A boolean list representing whether a value is
+#                     outside the process limits
+#                   - relative_to_mean: A list representing the relative value of an
+#                     element to mean
+#                   - close_to_limits: A boolean list representing whether a value is
+#                     close to a limit or not
+#                   - special_cause_flag: A boolean list representing whether a value
+#                     is a special cause or not'outside_limits' representing whether a
+#                    value is outside the process limits
+#
+#                 This test covers a single scenario reproducing the example from
+#                 https://www.england.nhs.uk/statistical-process-control-tool/
+#                 https://youtu.be/sWndrZ68Xww
+#                 More details in the test function docstring.
+#
+#                 TODO: Split the test into multiple tests by calculated field?
+#                 TODO: Include more scenarios
+#                 TODO: Test
+
+# CONTRIBUTORS:   Joan Ponsa, Craig R. Shenton
+# CONTACT:        [email protected]
+# CREATED:        18 Nov 2023
+# VERSION:        0.0.1
+
+# Imports
+# -------------------------------------------------------------------------
+# Python:
+import unittest
+
+# 3rd party:
+import pandas as pd
+import numpy as np
+
+# Local
+from nhspy_plotthedots.pandas_spc_calculations import pandas_spc_x_calc
+
+# Define tests
+# -------------------------------------------------------------------------
+
+
+class TestPandasSpcXCalc(unittest.TestCase):
+    def test_demo_case(self):
+        """
+        Test example reproduced from https://www.england.nhs.uk/statistical-process-control-tool/
+        demonstrating the calculation of the SPC. Notice the values are similar but not identical.
+        The resolution of the video was poor and I could not see the exact numbers.
+        Also, I subtracted 10 from the original values in order to obtain an upper process limit value.
+        """
+        values_arr = [89, 86, 75, 77, 81, 74, 72, 78, 76, 82, 77, 54, 66, 77, 86]
+        values_df = pd.DataFrame({"values": values_arr})
+        result_df = pandas_spc_x_calc(values_df, "values")
+
+        # test mean
+        expected_mean = 76.67
+        self.assertAlmostEqual(result_df["mean"].iloc[-1], expected_mean, places=2)
+
+        # test lpl and upl
+        expected_lpl = 57
+        expected_upl = 96
+        self.assertAlmostEqual(result_df["lpl"].iloc[-1], expected_lpl, places=0)
+        self.assertAlmostEqual(result_df["upl"].iloc[-1], expected_upl, places=0)
+
+        # test outside_limits
+        expected_outside_limits = [False] * 11 + [True] * 1 + [False] * 3
+        self.assertEqual(result_df["outside_limits"].tolist(), expected_outside_limits)
+
+        # test_relative_to_mean
+        expected_relative_to_mean = [
+            1.0 if value >= expected_mean else -1.0 for value in values_arr
+        ]
+        self.assertEqual(
+            result_df["relative_to_mean"].tolist(), expected_relative_to_mean
+        )
+
+        # test close_to_limits
+        expected_close_to_limits = [False] * 15
+        self.assertEqual(
+            result_df["close_to_limits"].tolist(), expected_close_to_limits
+        )
+
+        # test special_cause_flag
+        expected_special_cause_flag = [False] * 11 + [True] * 1 + [False] * 3
+        self.assertEqual(
+            result_df["special_cause_flag"].tolist(), expected_special_cause_flag
+        )
+
+    def test_fix_point_mean(self):
+        """
+        Test the mean calculation when the fix_point_mean is set.
+        """
+        values_df = pd.DataFrame({"values": np.arange(1, 11)})
+        result_df = pandas_spc_x_calc(values_df, "values", 5)
+        expected_mean = 3
+        self.assertAlmostEqual(result_df["mean"].iloc[-1], expected_mean, places=2)
+
+
+if __name__ == "__main__":
+    unittest.main()