
import unittest
import user_code
import ast
import re   
import importlib
import csv
import unittest
import pandas as pd
import numpy as np
import importlib

class TestTask(unittest.TestCase):
    def setUp(self):
        self.data = {
            'apple': [1, 0, 1, 0],
            'banana': [1, 1, 0, 0],
            'carrot': [0, 1, 1, 0],
            'donut': [0, 0, 1, 1],
        }
        self.user_item_matrix = pd.DataFrame(self.data, index=['alice', 'bob', 'carol', 'dave'])
        self.module_name = 'user_code'

    def test_returns_list(self):
        import user_code
        importlib.reload(user_code)
        result = user_code.recommend_products(self.user_item_matrix, 'bob', top_n=2)
        _dynamic_test(self, isinstance(result, list), "Function returns a list.", f"Expected list, got {type(result)}.")

    def test_recommendations_are_valid(self):
        import user_code
        importlib.reload(user_code)
        # Bob's vector: [0,1,1,0], most similar is Alice: [1,1,0,0]
        # Alice has apple (1), bob does not (0)
        result = user_code.recommend_products(self.user_item_matrix, 'bob', top_n=2)
        expected = ['apple']
        _dynamic_test(
            self,
            all(item in ['apple'] for item in result),
            "Recommendation is valid and contains only items the similar user has and target does not.",
            f"Recommendation contains unexpected items: {result}"
        )

    def test_top_n_limit(self):
        import user_code
        importlib.reload(user_code)
        # carol: [1,0,1,1], most similar user is alice: [1,1,0,0]
        # Alice has banana (1), carol does not (0)
        result = user_code.recommend_products(self.user_item_matrix, 'carol', top_n=1)
        _dynamic_test(
            self,
            isinstance(result, list) and len(result) <= 1 and (len(result) == 0 or result[0] == 'banana'),
            "Function respects top_n and returns correct recommendation for carol.",
            f"Expected ['banana'] or [], got {result}"
        )

    def test_no_similar_user(self):
        import user_code
        importlib.reload(user_code)
        # Add a new user with all zeros
        df = self.user_item_matrix.copy()
        df.loc['eve'] = [0, 0, 0, 0]
        result = user_code.recommend_products(df, 'eve', top_n=2)
        _dynamic_test(self, result == [], "Returns empty list if no similar user.", f"Expected [], got {result}")

    def test_no_new_items_to_recommend(self):
        import user_code
        importlib.reload(user_code)
        # Bob and dave: dave has only donut, bob has not. But similarity is zero with everyone except alice.
        # Add a user who has all the same items as alice
        df = self.user_item_matrix.copy()
        df.loc['erin'] = [1, 1, 0, 0]
        result = user_code.recommend_products(df, 'erin', top_n=2)
        _dynamic_test(
            self,
            result == [],
            "Returns empty list if most similar user has no new items to recommend.",
            f"Expected [], got {result}"
        )

    def test_handles_different_top_n(self):
        import user_code
        importlib.reload(user_code)
        # alice: [1,1,0,0], most similar user is bob: [0,1,1,0]
        # bob has carrot (1), alice does not (0)
        result = user_code.recommend_products(self.user_item_matrix, 'alice', top_n=5)
        _dynamic_test(
            self,
            result == ['carrot'],
            "Function returns all possible recommendations if top_n > available.",
            f"Expected ['carrot'], got {result}"
        )

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

def normalize_text(text):
    text = text.lower()
    text = re.sub(r"\\s{2,}", " ", text)
    text = re.sub(r"\\s*([,:?])\\s*", r"\\1 ", text)
    return text.strip()

def change_var(code: str, var_name: str, value: str) -> str:
    tree = ast.parse(code)
    lines = code.splitlines()
    changed = False
    # Collect all assignment nodes to modify
    assign_nodes = [
        (i, node)
        for i, node in enumerate(tree.body)
        if isinstance(node, ast.Assign)
        and any(isinstance(target, ast.Name) and target.id == var_name for target in node.targets)
    ]

    # If nothing to change, return unmodified code
    if not assign_nodes:
        return code

    # Perform replacements for all matching assignments (from last to first to not break line offsets)
    for i, node in reversed(assign_nodes):
        start_line = node.lineno - 1
        line = lines[start_line]
        indent = ' ' * (len(line) - len(line.lstrip()))
        lines[start_line] = f"{indent}{var_name} = {value}"
        next_line = len(lines)
        for next_node in tree.body[i+1:]:
            if hasattr(next_node, 'lineno'):
                next_line = next_node.lineno - 1
                break
        if next_line > start_line + 1:
            lines[start_line+1:next_line] = []
        changed = True

    return '\\n'.join(lines) if changed else code

if __name__ == "__main__":
    unittest.main()


test_main.py

Behersk de nødvendige færdigheder til at designe, implementere og skalere moderne anbefalingsmotorer og transaktionsbaserede associationsmodeller. Gå fra at analysere rå detailkøb ved hjælp af klassiske algoritmiske principper som Apriori og FP-Growth til at opbygge avancerede personaliseringssystemer via kollaborativ filtrering og matrixfaktorisering. Lær at udvikle datapipelines, der driver moderne digitale butiksfacader og øger brugerengagements-målinger.

Opdag, hvordan man finder skjulte relationer i massive transaktionsdatastrukturer. Lær de centrale matematiske rammer, som detailgiganter bruger til at optimere hyldeplacering, lagerstrukturer og produktbundlingsstrategier.

Hvad er den primære forskel mellem brugerbaseret og item-baseret kollaborativ filtrering?

**Hvilket mål er bedst til at måle lighed mellem brugerbedømmelsesvektorer?**

Hvordan forbedrer item-baseret filtrering skalerbarheden?

Hvornår vil du bruge Pearson-korrelation frem for cosinus-lighed?

Hvad er et nabolag i kollaborativ filtrering?

Tag din algoritmiske viden til næste niveau ved at håndtere store virksomhedstransaktionsdata uden hukommelsesnedbrud. Opnå ekspertise i højtydende komprimerede træer og lær, hvordan man eliminerer kommerciel støj fra regelmønstre.

Få praktisk erfaring med at opbygge anbefalingsmotorer, der driver moderne digitale platforme. Lær at se ud over produktgrupperinger for at kortlægge historiske brugeradfærd, vurderinger og delte forbrugsvaner.

Dyk ned i de matematiske grundprincipper for avancerede maskinlæringsmodeller til sparsomme datasæt. Lær, hvordan man afdækker latente egenskaber skjult i multidimensionelle brugerfeedbackdata for nøjagtigt at forudsige manglende interaktioner.

Challenge: Opbygning af en Brugerdrevet Kollaborativ Filtreringskomponent

Løsning