Python for Data Engineers Cheat Sheet

Files and CSV

import csv, io

f = io.StringIO("id,amount\n1,10\n2,20\n")          # use open(path, newline="") for real files
rows = list(csv.DictReader(f))
print(rows[0]["amount"])

JSON

import json

record = json.loads('{"id": 1, "tags": ["a", "b"]}')
print(json.dumps(record, sort_keys=True))

Dates and times

from datetime import date, datetime, timedelta, timezone

run_date = date(2026, 10, 1)
print(run_date.isoformat(), run_date + timedelta(days=1))
print(datetime(2026, 10, 1, 9, 30, tzinfo=timezone.utc).isoformat())

Store and compare timestamps in UTC; convert to local time only for display.

Collections

from collections import Counter, defaultdict, deque

totals = defaultdict(int)
for k, v in [("a", 1), ("b", 2), ("a", 3)]:
    totals[k] += v
print(dict(totals), Counter("abca").most_common(1), deque([1, 2, 3], maxlen=2))
print(list(dict.fromkeys(["b", "a", "b"])))          # dedupe, keep order

Generators and batching

from itertools import batched, islice

def read_ids(n):
    for i in range(n):
        yield i

print(list(islice(read_ids(10), 3)))
print(list(batched(read_ids(5), 2)))                  # Python 3.12+

Error handling

class LoadError(Exception):
    pass

try:
    try:
        int("x")
    except ValueError as exc:
        raise LoadError("bad amount in row 7") from exc
except LoadError as err:
    print(err, "| caused by", type(err.__cause__).__name__)

Catch specific exceptions, preserve the cause with raise ... from exc, and let unexpected errors fail the run.

Logging

import logging

logging.basicConfig(level=logging.INFO, format="%(asctime)s %(levelname)s %(name)s %(message)s")
log = logging.getLogger("orders")
log.info("loaded %d rows from %s", 120, "orders.csv")   # lazy formatting

Testing with pytest

def to_cents(amount: float) -> int:
    return int(round(amount * 100))

def test_to_cents():
    assert to_cents(19.99) == 1999

test_to_cents()

Run pytest -q. Put pure transformation logic in functions so it can be tested without I/O.