BUG: to_period() freq was not infered

[ShaharNaveh]

• closes DatetimeIndex.to_period with freq #33358
• tests added / passed
• passes black pandas
• passes git diff upstream/master -u -- "*.py" | flake8 --diff
• whatsnew entry

---

When running pytest -q --cache-clear pandas/tests/ -k "period" some tests in pandas/tests/indexing/ are failing, but when running pytest -q --cache-clear pandas/tests/indexing/ all the tests are passing (on my local machine), so I figured and debug from here, thoughts?

[jbrockmendel]

thoughts?

we cant rely on grepping to get all the relevant tests. i recommend pytest pandas/tests --skip-slow --skip-db, takes about 15 minutes for me.

Look at the failing tests, choose one and start executing it step-by-step in the REPL

off = "QS"
rng = pd.date_range("01-Jan-2012", periods=8, freq=off)

>>> prng = rng.to_period()   # <-- raises, so recurse into the `to_period` method

[jreback]

what is the warning from? this would not be set like this
use assert_produces_warning(None):
warnings.simplefiler....

if you really need to

[ShaharNaveh]

what is the warning from? this would not be set like this

The warning raises when a user is attempting to convert a date_range to period, if the date_range have a timezone it will drop it in the conversion, the warning let the user know about that:

pandas/pandas/core/arrays/datetimes.py

Lines 1085 to 1090 in 3cca07c

	if self.tz is not None:
	warnings.warn(
	"Converting to PeriodArray/Index representation "
	"will drop timezone information.",
	UserWarning,
	)

use assert_produces_warning(None):
warnings.simplefiler....

if you really need to

Sure! will fix.

[jreback]

i c, ok then

[jreback]

looks good, can you add a whatsnew note in 1.1 in the datetime section for bug fixes. ping on green.

[ShaharNaveh]

ping @jreback

[jreback]

thanks @MomIsBestFriend

[jbrockmendel]

@MomIsBestFriend why cant we just do res = freq here? (im trying to clean up usages of base_and_stride)