Pandas strftime(): Format Dates and Times

Format Pandas timestamps, date indexes and Series as text. Practise directives, missing dates, timezones and fiscal-period labels.

Show Table of Contents ↓

Format the original month-end date range 🔝

strftime() formats dates as text. Use DatetimeIndex.strftime(), Timestamp.strftime() or Series.dt.strftime(); there is no DataFrame.strftime(). The original October 2019 to February 2020 example uses the current ME month-end alias and unambiguous ISO input dates.

import pandas as pd
dates = pd.date_range(start="2019-10-20", end="2020-02-14", freq="ME")
formatted = dates.strftime("%d-%b-%Y")
print(formatted)
assert len(formatted) == 4
assert dates[0] == pd.Timestamp("2019-10-31")

Expected output

Index(['31-Oct-2019', '30-Nov-2019', '31-Dec-2019', '31-Jan-2020'], dtype='str')

Use a portable numeric format 🔝

Month and weekday names depend on the runtime locale. Numeric year-month-day formats are suitable when a stable output convention is needed.

iso = dates.strftime("%Y-%m-%d")
print(iso.tolist())
assert iso.tolist() == ["2019-10-31", "2019-11-30", "2019-12-31", "2020-01-31"]

Expected output

['2019-10-31', '2019-11-30', '2019-12-31', '2020-01-31']

Date and time directives 🔝

Directive	Meaning
`%a / %A`	Abbreviated / full weekday name; locale dependent.
`%w`	Weekday number, Sunday 0 to Saturday 6.
`%W / %U`	Week number with Monday / Sunday first; days before the first such weekday are week 00.
`%d / %j`	Day of month / day of year.
`%b / %B / %m`	Abbreviated month / full month / numeric month.
`%y / %Y`	Two-digit / full calendar year.
`%H / %I`	24-hour / 12-hour clock.
`%M / %S`	Minute / second.
`%p`	AM/PM; locale dependent.
`%f`	Microseconds, not a fiscal year for datetime values.
`%z / %Z`	UTC offset / timezone name.
`%x / %X / %c`	Locale-specific date / time / combined display.

The original fiscal-year and quarter entries need object-specific treatment: Period.strftime() supports %q, %f and %F; DatetimeIndex/Series datetime formatting does not provide the same fiscal-period semantics. Use the period example below. Directive support can also vary by operating system; prefer explicit numeric directives for exchange formats.

Format one timestamp and its components 🔝

Preserve the original hour, minute, second, weekday and day-of-year topics. This fixed synthetic timestamp keeps the example repeatable.

stamp = pd.Timestamp("2026-09-05 14:35:20.123456")
print(stamp.strftime("%Y-%m-%d %H:%M:%S.%f"))
print("Weekday/day of year:", stamp.strftime("%w %j"))
print("Week numbers:", stamp.strftime("%W %U"))
assert stamp.strftime("%H:%M:%S") == "14:35:20"

Expected output

2026-09-05 14:35:20.123456
Weekday/day of year: 6 248
Week numbers: 35 35

Format a DataFrame date column with dt 🔝

Parse source text first with to_datetime(). Keep the datetime column for sorting and calculations; add a separate display column.

table = pd.DataFrame({"date_text": ["2026-09-05", "invalid", None]})
table["date"] = pd.to_datetime(table["date_text"], format="%Y-%m-%d", errors="coerce")
table["display_date"] = table["date"].dt.strftime("%d/%m/%Y").astype("string")
print(table)
assert table["date"].isna().sum() == 2
assert table["display_date"].iloc[0] == "05/09/2026"

Expected output

    date_text       date display_date
0  2026-09-05 2026-09-05   05/09/2026
1     invalid        NaT         <NA>
2         NaN        NaT         <NA>

Choose an explicit missing-date display 🔝

Formatting does not repair failed dates. A display label can help readers while the source and missing datetime remain available for review.

display = table["display_date"].fillna("Needs review")
print(display.tolist())
assert display.tolist() == ["05/09/2026", "Needs review", "Needs review"]

Expected output

['05/09/2026', 'Needs review', 'Needs review']

Convert timezone before formatting 🔝

Format an aware timestamp after selecting the intended timezone. Text alone is insufficient for timezone calculations.

utc = pd.Timestamp("2026-09-05 09:00:00", tz="UTC")
local = utc.tz_convert("Asia/Kolkata")
print(local.strftime("%Y-%m-%d %H:%M %z %Z"))
assert local.strftime("%H:%M %z") == "14:30 +0530"

Expected output

2026-09-05 14:30 +0530 IST

Format fiscal periods and quarters 🔝

Keep the original fiscal-year and quarter concepts using the appropriate Period object. Q-MAR describes a fiscal year ending in March. These period aliases differ from timestamp month/quarter-end offset aliases.

period = pd.Period("2026-05", freq="Q-MAR")
print(period.strftime("Fiscal %F, short year %f, quarter %q"))
assert period.strftime("%F Q%q") == "2027 Q1"

Expected output

Fiscal 2027, short year 27, quarter 1

Common mistakes 🔝

Do not call DataFrame.strftime(), confuse %m (month) with %M (minute), or replace datetime columns with display strings before calculations. %b, %A, %p and local %x/%X/%c depend on locale. Do not infer fiscal years from datetime %f. See period_range() for period sequences and date_range() for timestamp sequences.

Exercise: create report labels 🔝

Create two daily timestamps starting 2026-09-05, format them as YYYYMMDD, and verify that the underlying dates are unchanged. Continue with Date and Time exercises.

Exercise solution 🔝

Keep the date sequence separate from its display labels.

report_dates = pd.date_range("2026-09-05", periods=2, freq="D")
labels = report_dates.strftime("%Y%m%d")
print(labels.tolist())
assert labels.tolist() == ["20260905", "20260906"]
assert report_dates[0] == pd.Timestamp("2026-09-05")

Expected output

['20260905', '20260906']

Practice in Google Colab 🔝

Open in Google Colab View on GitHub

Run examples in order, inspect formatted dates and try the exercise. Save a copy to keep edits.

Reference: Pandas strftime() documentation.

Subhendu Mohapatra

Author & Instructor at plus2net

I write and maintain practical tutorials on Python, PHP, SQL, JavaScript, HTML, jQuery, and web development at plus2net. The tutorials focus on clear explanations, working examples, and code that readers can test and adapt while learning.

About the author Watch on YouTube

Connect:

Subscribe to our YouTube Channel here