Stop rebuilding your CSV totals: a small Python script

Aggregate a three-row CSV without changing the source file

A worked example with fictional data, not a customer result.

books 1,200
software 600
books 800

books 2,000
software 600

category,amount
books,2000
software,600

Console: 2600 total; 2 categories

The books rows add up to 2,000 and software adds up to 600, giving 2,600 in total. Three input rows become two categories. Use non-negative integers in one consistent unit. Unclosed quotes, extra or missing columns, negative amounts, decimals, and an input with no records stop the script before it creates a result.

Try it with this input

Start with this input

category,amount
books,1200
software,600
books,800

Follow these steps

1. Check that Python 3 is available. On Windows, open PowerShell and run python --version or py --version. On macOS/Linux, open a terminal and run python3 --version. Use the command that reports Python 3.x; no package installation is needed.
2. Create an empty working folder. Save the input as weekly.csv and the code as summarize.py, both as plain text in UTF-8. Check that the names do not end in .txt. Amounts must be non-negative integers in the same unit.
3. In Windows PowerShell or a macOS/Linux terminal, run cd "full path to your working folder". Check the location with Get-Location in PowerShell or pwd on macOS/Linux. Then run dir in PowerShell or ls on macOS/Linux and confirm that both weekly.csv and summarize.py are listed.
4. Run the script with the command you checked: python summarize.py or py summarize.py on Windows, or python3 summarize.py on macOS/Linux.
5. Compare totals.csv and the console total with the expected result: 2,600 and two categories. If the script stops, use the troubleshooting notes below before using any result.
6. For another run, copy weekly.csv and summarize.py into a new empty folder. An existing totals.csv makes the script stop rather than overwrite it.

Runnable example

import csv
from collections import defaultdict
from pathlib import Path

totals = defaultdict(int)
with Path("weekly.csv").open(encoding="utf-8-sig", newline="") as source:
    reader = csv.DictReader(source, strict=True)
    if len(reader.fieldnames or []) != 2 or set(reader.fieldnames) != {"category", "amount"}:
        raise ValueError("Expected category and amount columns")
    for line, row in enumerate(reader, start=2):
        if None in row or any(value is None for value in row.values()):
            raise ValueError(f"Wrong column count on line {line}")
        category = row["category"].strip()
        raw_amount = row["amount"].strip()
        if not category or category[0] in "=+-@" or not raw_amount.isascii() or not raw_amount.isdigit():
            raise ValueError(f"Invalid category or non-negative integer amount on line {line}")
        totals[category] += int(raw_amount)

if not totals:
    raise ValueError("No data rows; check the input export")

# Exclusive creation: an existing result is never overwritten.
with Path("totals.csv").open("x", encoding="utf-8", newline="") as target:
    writer = csv.writer(target)
    writer.writerow(["category", "amount"])
    writer.writerows(sorted(totals.items()))
print(f"{sum(totals.values())} total; {len(totals)} categories")

If the example stops

Check before using the result

• Work on a copy and confirm that weekly.csv has not changed
• Check both column names, encoding, and a consistent unit for every amount
• Reconcile the input total of 2,600 with the output total of 2,600
• Compare total setup, runtime, review, and repair time against the manual task

Would another approach fit better?

Scroll the table sideways to read all columns.

Next decision

Try three copies of the same format and record the results and effort. Stay with the standard library while it meets the need. Compare a paid tool only after identifying a specific missing transformation, its maintenance cost, and the total price.

No paid product is required for this example. This page does not link to a product offer.