Performance
datetime operations can be slow at scale. Cache timezone objects, avoid parsing in loops, and use pandas for bulk operations.
Python
Advanced
4 min read
Datetime Performance Tips
Example
from zoneinfo import ZoneInfo
from datetime import datetime
# Cache timezone object (reuse, not re-instantiate)
UTC = ZoneInfo("UTC")
NY = ZoneInfo("America/New_York")
# Bulk conversion with pandas (much faster than loop)
import pandas as pd
ts_series = pd.Series([1718456400, 1718542800, 1718629200])
datetimes = pd.to_datetime(ts_series, unit="s", utc=True)
# Avoid strptime in tight loops — parse once, add timedelta
base = datetime(2024, 1, 1, tzinfo=UTC)
dates = [base + timedelta(i) for i in range(365)] Pro Tip
pd.to_datetime() on a Series is ~100x faster than a list comprehension of datetime.fromtimestamp().