Bij het werken met grote datasets is het vaak nodig om aggregaties uit te voeren zonder het volledige bestand in het geheugen te laden. Een veelvoorkomende taak is het optellen van de waarden van een specifieke kolom in een zeer groot CSV-bestand. Omdat het bestand mogelijk niet in het geheugen past, kun je het verwerken in beheersbare stukken met behulp van de `pandas`-functie `read_csv()` met de parameter `chunksize`.

Voor elk stuk bereken je de som van de gewenste kolom en vervolgens aggregeer je deze gedeeltelijke totalen om het uiteindelijke totaal te verkrijgen. Deze aanpak is efficiënt en schaalbaar, waardoor je bestanden van vrijwel elke grootte kunt verwerken zolang elk stuk in het geheugen past.


import unittest
import user_code
import ast
import re   
import importlib
import csv
import unittest
import importlib
import os
import csv
import pandas as pd

class TestTask(unittest.TestCase):
    def setUp(self):
        # Create test CSV files for different scenarios
        self.one_chunk_file = "test_one_chunk.csv"
        self.multi_chunk_file = "test_multi_chunk.csv"
        self.float_file = "test_float.csv"
        self.empty_col_file = "test_empty_col.csv"
        self.no_col_file = "test_no_col.csv"

        # One chunk: 3 rows
        with open(self.one_chunk_file, "w", newline="") as f:
            writer = csv.writer(f)
            writer.writerow(["amount", "desc"])
            writer.writerow([10, "a"])
            writer.writerow([20, "b"])
            writer.writerow([30, "c"])
        # Multi chunk: 6 rows
        with open(self.multi_chunk_file, "w", newline="") as f:
            writer = csv.writer(f)
            writer.writerow(["value", "desc"])
            for i in range(1, 7):
                writer.writerow([i, f"row{i}"])
        # Float values
        with open(self.float_file, "w", newline="") as f:
            writer = csv.writer(f)
            writer.writerow(["price", "desc"])
            writer.writerow([1.5, "x"])
            writer.writerow([2.5, "y"])
            writer.writerow([3.0, "z"])
        # Empty column
        with open(self.empty_col_file, "w", newline="") as f:
            writer = csv.writer(f)
            writer.writerow(["col1", "col2"])
            writer.writerow(["", "a"])
            writer.writerow(["", "b"])
        # No such column
        with open(self.no_col_file, "w", newline="") as f:
            writer = csv.writer(f)
            writer.writerow(["foo", "bar"])
            writer.writerow([1, 2])
            writer.writerow([3, 4])

    def tearDown(self):
        for f in [self.one_chunk_file, self.multi_chunk_file, self.float_file, self.empty_col_file, self.no_col_file]:
            if os.path.exists(f):
                os.remove(f)

    def test_sum_one_chunk(self):
        import user_code
        importlib.reload(user_code)
        result = user_code.sum_column_in_chunks(self.one_chunk_file, "amount", 100)
        _dynamic_test(
            self,
            result == 60,
            "Returns correct sum when file fits in one chunk.",
            f"Expected 60, got {result} for one chunk file."
        )

    def test_sum_multiple_chunks(self):
        import user_code
        importlib.reload(user_code)
        # chunksize=2, so 3 chunks
        result = user_code.sum_column_in_chunks(self.multi_chunk_file, "value", 2)
        _dynamic_test(
            self,
            result == 21,
            "Returns correct sum when file is read in multiple chunks.",
            f"Expected 21, got {result} for multi-chunk file."
        )

    def test_sum_float_column(self):
        import user_code
        importlib.reload(user_code)
        result = user_code.sum_column_in_chunks(self.float_file, "price", 2)
        _dynamic_test(
            self,
            abs(result - 7.0) < 1e-8,
            "Returns correct sum for float column.",
            f"Expected 7.0, got {result} for float column."
        )

    def test_sum_empty_column(self):
        import user_code
        importlib.reload(user_code)
        result = user_code.sum_column_in_chunks(self.empty_col_file, "col1", 1)
        _dynamic_test(
            self,
            result == 0,
            "Returns 0 for column with no values.",
            f"Expected 0, got {result} for empty column."
        )

    def test_sum_column_not_exist(self):
        import user_code
        importlib.reload(user_code)
        result = user_code.sum_column_in_chunks(self.no_col_file, "notacol", 1)
        _dynamic_test(
            self,
            result == 0,
            "Returns 0 if column does not exist.",
            f"Expected 0, got {result} for missing column."
        )

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

Een praktische, hands-on cursus voor aspirant-data scientists die klaar zijn om echte uitdagingen met grote datasets aan te gaan. Leer efficiënt grote datasets verwerken, samplen en analyseren met Python en essentiële libraries. Elke sectie bevat boeiende video-uitleg en interactieve uitdagingen om je expertise op te bouwen.

Leer fundamentele strategieën voor het omgaan met datasets die te groot zijn om in het geheugen te passen, waaronder chunking- en streamingtechnieken.

Ontdek methoden om grote datasets te balanceren en te bemonsteren, waaronder oversampling en undersampling.

Leer hoe u de polars-bibliotheek gebruikt voor snelle, geheugen-efficiënte gegevensverwerking.

Challenge: Gegevensaggregatie in Delen

Oplossing