import unittest
import numpy as np
import user_code  # ÐºÐ¾Ð´ ÑÑÑÐ´ÐµÐ½ÑÐ°

def _dynamic(tc, ok, ok_msg, fail_msg):
    tc._testMethodName = ok_msg if ok else fail_msg
    (tc.assertTrue if ok else tc.fail)(tc._testMethodName)

ATOL = 1e-10
RTOL = 1e-9

class TestQualityControlProb(unittest.TestCase):

    def test_variables_exist_and_types(self):
        reasons = []

        # Scalars & counts
        total_rods = getattr(user_code, "total_rods", None)
        defective = getattr(user_code, "defective", None)
        long_ = getattr(user_code, "long", None)
        defective_and_long = getattr(user_code, "defective_and_long", None)

        # Probabilities
        P_D_or_L = getattr(user_code, "P_D_or_L", None)
        P_D_given_L = getattr(user_code, "P_D_given_L", None)

        # Sampling params & outputs
        population_mean = getattr(user_code, "population_mean", None)
        population_std = getattr(user_code, "population_std", None)
        sample_size = getattr(user_code, "sample_size", None)
        sample_lengths = getattr(user_code, "sample_lengths", None)
        sample_mean = getattr(user_code, "sample_mean", None)
        sample_variance = getattr(user_code, "sample_variance", None)
        sample_std = getattr(user_code, "sample_std", None)

        # Basic presence & type checks
        if not isinstance(total_rods, (int, np.integer)):
            reasons.append("total_rods must be int")
        if not isinstance(defective, (int, np.integer)):
            reasons.append("defective must be int")
        if not isinstance(long_, (int, np.integer)):
            reasons.append("long must be int")
        if not isinstance(defective_and_long, (int, np.integer)):
            reasons.append("defective_and_long must be int")

        if not isinstance(P_D_or_L, (float, np.floating)):
            reasons.append("P_D_or_L must be float")
        if not isinstance(P_D_given_L, (float, np.floating)):
            reasons.append("P_D_given_L must be float")

        if not isinstance(population_mean, (int, float, np.integer, np.floating)):
            reasons.append("population_mean must be numeric")
        if not isinstance(population_std, (int, float, np.integer, np.floating)):
            reasons.append("population_std must be numeric")
        if not isinstance(sample_size, (int, np.integer)):
            reasons.append("sample_size must be int")

        if not (isinstance(sample_lengths, np.ndarray) and sample_lengths.shape == (sample_size,)):
            reasons.append(f"sample_lengths must be ndarray of shape ({sample_size},)")
        if not isinstance(sample_mean, (float, np.floating)):
            reasons.append("sample_mean must be float")
        if not isinstance(sample_variance, (float, np.floating)):
            reasons.append("sample_variance must be float")
        if not isinstance(sample_std, (float, np.floating)):
            reasons.append("sample_std must be float")

        _dynamic(self,
                 len(reasons) == 0,
                 "PASS: variables exist with correct types/shapes",
                 "FAIL: " + " | ".join(reasons) if reasons else "FAIL")

    def test_probabilities_correct(self):
        reasons = []

        total_rods = user_code.total_rods
        defective = user_code.defective
        long_ = user_code.long
        defective_and_long = user_code.defective_and_long

        expected_union = (defective/total_rods) + (long_/total_rods) - (defective_and_long/total_rods)
        expected_cond = (defective_and_long/total_rods) / (long_/total_rods)

        if not np.isclose(user_code.P_D_or_L, expected_union, atol=ATOL, rtol=RTOL):
            reasons.append(f"P(D âª L) incorrect: expected {expected_union}, got {user_code.P_D_or_L}")
        if not np.isclose(user_code.P_D_given_L, expected_cond, atol=ATOL, rtol=RTOL):
            reasons.append(f"P(D|L) incorrect: expected {expected_cond}, got {user_code.P_D_given_L}")

        _dynamic(self,
                 len(reasons) == 0,
                 "PASS: probabilities (union & conditional) are correct",
                 "FAIL: " + " | ".join(reasons) if reasons else "FAIL")

    def test_sampling_reproducible_and_stats(self):
        reasons = []

        # Recompute expected sample deterministically with the same seed & params
        population_mean = user_code.population_mean
        population_std = user_code.population_std
        sample_size = user_code.sample_size

        np.random.seed(0)
        expected_sample = np.random.normal(loc=population_mean, scale=population_std, size=sample_size)
        expected_mean = np.mean(expected_sample)
        expected_var = np.var(expected_sample, ddof=1)
        expected_std = np.std(expected_sample, ddof=1)

        # Compare arrays & stats
        if not (isinstance(user_code.sample_lengths, np.ndarray) and user_code.sample_lengths.shape == expected_sample.shape):
            reasons.append("sample_lengths shape mismatch")
        else:
            if not np.allclose(user_code.sample_lengths, expected_sample, atol=1e-12, rtol=1e-12):
                reasons.append("sample_lengths not reproducible with seed=0 and given parameters")

        if not np.isclose(user_code.sample_mean, expected_mean, atol=ATOL, rtol=RTOL):
            reasons.append(f"sample_mean incorrect: expected {expected_mean}, got {user_code.sample_mean}")
        if not np.isclose(user_code.sample_variance, expected_var, atol=ATOL, rtol=RTOL):
            reasons.append(f"sample_variance incorrect (use ddof=1): expected {expected_var}, got {user_code.sample_variance}")
        if not np.isclose(user_code.sample_std, expected_std, atol=ATOL, rtol=RTOL):
            reasons.append(f"sample_std incorrect (use ddof=1): expected {expected_std}, got {user_code.sample_std}")

        _dynamic(self,
                 len(reasons) == 0,
                 "PASS: sampling is reproducible & stats (mean/var/std) correct",
                 "FAIL: " + " | ".join(reasons) if reasons else "FAIL")


test_code.py

Beherrschung der mathematischen Grundlagen, die für Data Science unerlässlich sind. Untersuchung zentraler Konzepte in Funktionen, Analysis, Linearer Algebra, Wahrscheinlichkeit und Dimensionsreduktion. Aufbau sowohl theoretischer Kenntnisse als auch praktischer Programmiererfahrung zur Stärkung der Fähigkeiten in Datenanalyse, Modellierung komplexer Systeme und Anwendung fortgeschrittener Methoden im maschinellen Lernen.

Erkunden Sie die Grundlagen mathematischer Funktionen. Verschiedene Typen algebraischer und transzendenter Funktionen, deren Eigenschaften sowie deren Implementierung in Python zur Lösung praxisnaher Probleme.

Beherrschen Sie die Konzepte von Mengen und Reihen, von grundlegenden Operationen bis zu praktischen Anwendungen. Sammeln Sie praktische Erfahrungen bei der Implementierung von Mengenoperationen sowie beim Arbeiten mit arithmetischen und geometrischen Reihen in Python.

Fundiertes Verständnis von Grenzwerten, Ableitungen, Integralen und partiellen Ableitungen. Verbindung von Theorie und Praxis durch Implementierung dieser Konzepte in Python und Anwendung auf Optimierung mittels Gradientenabstieg.

Fundierte Kenntnisse über Vektoren, Matrizen und Transformationen. Vermittlung von Zerlegungsmethoden und Eigenwertanalyse, mit Vertiefung der Konzepte durch Python-Programmierungsaufgaben und praxisnahe Anwendungen in der Datenwissenschaft.

Vertiefung in Wahrscheinlichkeitstheorie und Statistik. Untersuchung von bedingter Wahrscheinlichkeit, Satz von Bayes und statistischen Kennzahlen. Umsetzung zentraler Konzepte in Python, Simulation von Verteilungen sowie Festigung der Kenntnisse durch Herausforderungen und Quizfragen.

Herausforderung: Wahrscheinlichkeitsanalyse in der Qualitätskontrolle

Lösung

Awesome!

Herausforderung: Wahrscheinlichkeitsanalyse in der Qualitätskontrolle

Lösung