Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Handle time normalization for nonexistent and ambiguous times #2133

Open
wants to merge 12 commits into
base: main
Choose a base branch
from
4 changes: 3 additions & 1 deletion docs/sphinx/source/whatsnew/v0.11.1.rst
Original file line number Diff line number Diff line change
Expand Up @@ -26,7 +26,8 @@ Enhancements

Bug fixes
~~~~~~~~~

* Handle DST transitions that happen at midnight in :py:func:`pvlib.solarposition.hour_angle`
(:issue:`2132` :pull:`2133`)

Testing
~~~~~~~
Expand Down Expand Up @@ -55,4 +56,5 @@ Contributors
* Leonardo Micheli (:ghuser:`lmicheli`)
* Echedey Luis (:ghuser:`echedey-ls`)
* Rajiv Daxini (:ghuser:`RDaxini`)
* Scott Nelson (:ghuser:`scttnlsn`)
* Jose Meza (:ghuser:`JoseMezaMendieta`)
17 changes: 16 additions & 1 deletion pvlib/solarposition.py
Original file line number Diff line number Diff line change
Expand Up @@ -1360,6 +1360,11 @@ def hour_angle(times, longitude, equation_of_time):
times : :class:`pandas.DatetimeIndex`
Corresponding timestamps, must be localized to the timezone for the
``longitude``.

A `pytz.exceptions.AmbiguousTimeError` could be raised if any of the
given times are on a day where the local daylight savings transition happened
at midnight. If you're working with such a timezone, consider converting to
a non-DST timezone (i.e. GMT-4) before calling this function.
scttnlsn marked this conversation as resolved.
Show resolved Hide resolved
longitude : numeric
Longitude in degrees
equation_of_time : numeric
Expand Down Expand Up @@ -1391,7 +1396,17 @@ def hour_angle(times, longitude, equation_of_time):
times = times.tz_localize('utc')
tzs = np.array([ts.utcoffset().total_seconds() for ts in times]) / 3600

hrs_minus_tzs = (times - times.normalize()) / pd.Timedelta('1h') - tzs
# Some timezones have a DST shift at midnight:
# 11:59pm -> 1:00am - results in a nonexistent midnight
# 12:59am -> 12:00am - results in an ambiguous midnight
# We remove the timezone before normalizing for this reason.
naive_normalized_times = times.tz_localize(None).normalize()

# Use Pandas functionality for shifting nonexistent times forward
normalized_times = naive_normalized_times.tz_localize(
times.tz, nonexistent='shift_forward', ambiguous='raise')
Comment on lines +1405 to +1407
Copy link
Contributor

@yhkee0404 yhkee0404 Sep 12, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What about this with the modification of def hour_angle(times, longitude, equation_of_time) into def hour_angle(times, longitude, equation_of_time, nonexistent='shift_forward', ambiguous='raise') to give users a choice to control the tz_localize behavior if they want rather than just receiving an error and being blocked?

Suggested change
# Use Pandas functionality for shifting nonexistent times forward
normalized_times = naive_normalized_times.tz_localize(
times.tz, nonexistent='shift_forward', ambiguous='raise')
# Use Pandas functionality for shifting nonexistent times forward
normalized_times = naive_normalized_times.tz_localize(
times.tz, nonexistent=nonexistent, ambiguous=ambiguous)

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@scttnlsn could you comment on this?

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The user would not necessarily be blocked if the exception is raised. They'd just need to convert the given times to a non-DST timezone ahead of calling hour_angle. The documentation is updated to mention this.

infer won't really work as mentioned here: #2133 (comment) The other options are to raise (current approach) or to have NaNs returned where the times are ambiguous (the user still likely needs to deal with that case).

I'm personally of the opinion that forcing the user to convert to a non-DST timezone is a fine approach but I could understand someone might want to have NaNs instead.


hrs_minus_tzs = (times - normalized_times) / pd.Timedelta('1h') - tzs

# ensure array return instead of a version-dependent pandas <T>Index
return np.asarray(
Expand Down
33 changes: 33 additions & 0 deletions pvlib/tests/test_solarposition.py
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@
from .conftest import assert_frame_equal, assert_series_equal
from numpy.testing import assert_allclose
import pytest
import pytz

from pvlib.location import Location
from pvlib import solarposition, spa
Expand Down Expand Up @@ -673,6 +674,38 @@ def test_hour_angle():
assert np.allclose(hours, expected)


def test_hour_angle_with_tricky_timezones():
# GH 2132
# tests timezones that have a DST shift at midnight
cwhanse marked this conversation as resolved.
Show resolved Hide resolved
scttnlsn marked this conversation as resolved.
Show resolved Hide resolved

eot = np.array([-3.935172, -4.117227, -4.026295, -4.026295])

longitude = 70.6693
times = pd.DatetimeIndex([
'2014-09-06 23:00:00',
'2014-09-07 00:00:00',
'2014-09-07 01:00:00',
'2014-09-07 02:00:00',
]).tz_localize('America/Santiago', nonexistent='shift_forward')

with pytest.raises(pytz.exceptions.NonExistentTimeError):
times.normalize()

# should not raise `pytz.exceptions.NonExistentTimeError`
solarposition.hour_angle(times, longitude, eot)

longitude = 82.3666
times = pd.DatetimeIndex([
'2014-11-01 23:00:00',
'2014-11-02 00:00:00',
'2014-11-02 01:00:00',
'2014-11-02 02:00:00',
]).tz_localize('America/Havana', ambiguous=[True, True, False, False])

with pytest.raises(pytz.exceptions.AmbiguousTimeError):
solarposition.hour_angle(times, longitude, eot)


def test_sun_rise_set_transit_geometric(expected_rise_set_spa, golden_mst):
"""Test geometric calculations for sunrise, sunset, and transit times"""
times = expected_rise_set_spa.index
Expand Down