import unittest
import importlib
import pandas as pd
import seaborn as sns

def _dynamic_test(test_case, condition, success_msg, failure_msg):
    if condition:
        test_case._testMethodName = success_msg
        test_case.assertTrue(True, success_msg)
    else:
        test_case._testMethodName = failure_msg
        test_case.fail(failure_msg)

class TestTitanicCleaning(unittest.TestCase):
    @classmethod
    def setUpClass(cls):
        """Import user's solution and prepare reference data for testing."""
        cls.user_code = importlib.import_module("user_code")
        cls.cleaned = cls.user_code.cleaned_data

        # Load base dataset to compute IQR bounds for reference
        base = sns.load_dataset("titanic")
        base["age"].fillna(base["age"].mean(), inplace=True)
        base["embarked"].fillna(base["embarked"].mode()[0], inplace=True)
        base.drop_duplicates(inplace=True)

        Q1 = base["fare"].quantile(0.25)
        Q3 = base["fare"].quantile(0.75)
        IQR = Q3 - Q1
        cls.lower_bound = Q1 - 1.5 * IQR
        cls.upper_bound = Q3 + 1.5 * IQR

    def test_no_missing_values(self):
        """Ensure missing values in 'age' and 'embarked' are handled."""
        df = self.cleaned
        condition = (
            not df["age"].isnull().any() and
            not df["embarked"].isnull().any()
        )
        _dynamic_test(
            self,
            condition,
            "No missing values remain in 'age' or 'embarked'.",
            "Missing values still exist in 'age' or 'embarked'."
        )

    def test_no_duplicates(self):
        """Ensure duplicate rows are removed."""
        df = self.cleaned
        condition = len(df) == len(df.drop_duplicates())
        _dynamic_test(
            self,
            condition,
            "Duplicate rows successfully removed.",
            "Duplicate rows still exist in the dataset."
        )

    def test_outlier_removal(self):
        """Verify that outliers in 'fare' were filtered using IQR."""
        df = self.cleaned.copy()
        fares = df["fare"].dropna()
        eps = 1e-6
        within = fares.between(self.lower_bound - eps, self.upper_bound + eps)
        condition = within.all()
        _dynamic_test(
            self,
            condition,
            "Outliers in 'fare' handled correctly using IQR.",
            "Some extreme 'fare' values were not properly filtered."
        )

    def test_columns_exist(self):
        """Ensure required columns exist in the cleaned dataset."""
        required = {"age", "embarked", "fare"}
        df_cols = set(self.cleaned.columns)
        condition = required.issubset(df_cols)
        _dynamic_test(
            self,
            condition,
            "Cleaned dataset includes essential columns.",
            "One or more essential columns are missing in the cleaned dataset."
        )

if __name__ == "__main__":
    unittest.main(argv=["first-arg-is-ignored"], exit=False)


test_code.py

Covers essential techniques for preparing raw data for supervised learning. Focuses on handling missing values, encoding categorical features, scaling and transforming numerical data, and creating meaningful features that improve model performance and reliability.


Challenge: Data Cleaning

Løsning