shaare.it

TypeError: Conflict between tz_localize and tz_convert (timezone aware vs naive)

7 Dec 2025

2 min read

python pandas errors

TypeError: Already tz-aware / Cannot convert tz-naive index to timezone-aware

$ python - <<'PY'
import pandas as pd

idx = pd.date_range('2020-01-01', periods=2, tz='UTC')
s = pd.Series([1, 2], index=idx)
# Trying to localize an index that's already timezone-aware
s.tz_localize('Europe/Paris')
PY
Traceback (most recent call last):
  File "<string>", line 1, in <module>
TypeError: Already tz-aware, use tz_convert to convert timezone

And the converse:

$ python - <<'PY'
import pandas as pd

idx = pd.date_range('2020-01-01', periods=2)  # naive
s = pd.Series([1, 2], index=idx)
# Trying to convert a naive index
s.tz_convert('UTC')
PY
Traceback (most recent call last):
  File "<string>", line 1, in <module>
TypeError: Cannot convert tz-naive timestamps, use tz_localize to localize

Why this happens

  • tz_localize attaches a timezone to naive datetimes. Calling it on an already timezone-aware index is a mistake; pandas will raise a TypeError telling you to use tz_convert instead.
  • tz_convert changes the timezone of timezone-aware timestamps. If you call it on a naive index, pandas doesn’t know which timezone to convert from and will raise a TypeError telling you to tz_localize first.

Fix

  • If your index is timezone-aware and you want to switch zones, use tz_convert.
  • If your index is tz-naive and you want to interpret it in a timezone, use tz_localize.
  • Check df.index.tz (or s.index.tz) to see whether the index is timezone-aware (returns a timezone) or naive (None).

Wrong code

# tz-aware index - this is wrong
df.index = pd.date_range('2020-01-01', periods=2, tz='UTC')
df.tz_localize('Europe/Paris')

# tz-naive index - wrong to use tz_convert
df.index = pd.date_range('2020-01-01', periods=2)
df.tz_convert('UTC')

Fixed code

# Convert already aware index to different timezone
df.index = pd.date_range('2020-01-01', periods=2, tz='UTC')
df = df.tz_convert('Europe/Paris')

# Localize naive index first, then convert if needed
df.index = pd.date_range('2020-01-01', periods=2)
df = df.tz_localize('UTC').tz_convert('Europe/Paris')

Always inspect df.index.tz before calling tz_localize or tz_convert. Localizing or converting incorrectly can silently produce wrong results if not done as intended.