Add tests for time conversions in tools package

docs/sphinx/source/whatsnew/v0.11.3.rst

@@ -12,12 +12,21 @@ Enhancements
 ~~~~~~~~~~~~
 
 
+Bug Fixes
+~~~~~~~~~
+* Ensure proper tz and pytz types in pvlib.location.Location.
+  (:issue:`2340`, :pull:`2341`)
+
+
 Documentation
 ~~~~~~~~~~~~~
 
 
 Testing
 ~~~~~~~
+* Add tests for timezone types in pvlib.location.Location.
+  (:issue:`2340`, :pull:`2341`)
+* Add tests for time conversions in pvlib.tools. (:issue:`2340`, :pull:`2341`)
 
 
 Requirements
@@ -26,5 +35,4 @@ Requirements
 
 Contributors
 ~~~~~~~~~~~~
-
-
+* Mark Campanellli (:ghuser:`markcampanelli`)

pvlib/location.py

@@ -18,13 +18,15 @@
 class Location:
     """
     Location objects are convenient containers for latitude, longitude,
-    timezone, and altitude data associated with a particular
-    geographic location. You can also assign a name to a location object.
+    time zone, and altitude data associated with a particular geographic
+    location. You can also assign a name to a location object.
 
-    Location objects have two timezone attributes:
+    Location objects have two time-zone attributes, either of which can be
+    individually changed after the Location object has been instantiated and
+    the other will stay in sync:
 
-        * ``tz`` is a IANA timezone string.
-        * ``pytz`` is a pytz timezone object.
+        * ``tz`` is a IANA time-zone string.
+        * ``pytz`` is a pytz time-zone object.
 
     Location objects support the print method.
 
@@ -38,12 +40,16 @@ class Location:
         Positive is east of the prime meridian.
         Use decimal degrees notation.
 
-    tz : str, int, float, or pytz.timezone, default 'UTC'.
-        See
-        http://en.wikipedia.org/wiki/List_of_tz_database_time_zones
-        for a list of valid time zones.
-        pytz.timezone objects will be converted to strings.
-        ints and floats must be in hours from UTC.
+    tz : time zone as str, int, float, or datetime.tzinfo (inc. subclasses
+        from the pytz and zoneinfo packages), default 'UTC'.
+        See http://en.wikipedia.org/wiki/List_of_tz_database_time_zones for a
+        list of valid name strings for IANA time zones.
+        ints and floats must be non-fractional N-hour offsets from UTC, which
+        are converted to the 'Etc/GMT-N' format (note limited range of N and
+        its conventional sign change).
+        Raises TypeError for time zone conversion issues or
+        pytz.exceptions.UnknownTimeZoneError when (stringified) time zone is
+        not recognized by pytz.timezone.
 
     altitude : float, optional
         Altitude from sea level in meters.
@@ -59,38 +65,54 @@ class Location:
     pvlib.pvsystem.PVSystem
     """
 
-    def __init__(self, latitude, longitude, tz='UTC', altitude=None,
-                 name=None):
-
+    def __init__(
+        self, latitude, longitude, tz='UTC', altitude=None, name=None
+    ):
         self.latitude = latitude
         self.longitude = longitude
-
-        if isinstance(tz, str):
-            self.tz = tz
-            self.pytz = pytz.timezone(tz)
-        elif isinstance(tz, datetime.timezone):
-            self.tz = 'UTC'
-            self.pytz = pytz.UTC
-        elif isinstance(tz, datetime.tzinfo):
-            self.tz = tz.zone
-            self.pytz = tz
-        elif isinstance(tz, (int, float)):
-            self.tz = tz
-            self.pytz = pytz.FixedOffset(tz*60)
-        else:
-            raise TypeError('Invalid tz specification')
+        self.tz = tz
 
         if altitude is None:
             altitude = lookup_altitude(latitude, longitude)
 
         self.altitude = altitude
-
         self.name = name
 
     def __repr__(self):
         attrs = ['name', 'latitude', 'longitude', 'altitude', 'tz']
+        # Use None as getattr default in case __repr__ is called during
+        # initialization before all attributes have been assigned.
         return ('Location: \n  ' + '\n  '.join(
-            f'{attr}: {getattr(self, attr)}' for attr in attrs))
+            f'{attr}: {getattr(self, attr, None)}' for attr in attrs))
+
+    @property
+    def tz(self):
+        # self.pytz holds the single source of time-zone truth.
+        return self.pytz.zone
+
+    @tz.setter
+    def tz(self, tz_):
+        if isinstance(tz_, str):
+            self.pytz = pytz.timezone(tz_)
+        elif isinstance(tz_, int):
+            self.pytz = pytz.timezone(f"Etc/GMT{-tz_:+d}")
+        elif isinstance(tz_, float):
+            if tz_ % 1 != 0:
+                raise TypeError(
+                    "floating point tz does not have zero fractional part: "
+                    f"{tz_}"
+                )
+
+            self.pytz = pytz.timezone(f"Etc/GMT{-int(tz_):+d}")
+        elif isinstance(tz_, datetime.tzinfo):
+            # Includes time zones generated by pytz and zoneinfo packages.
+            self.pytz = pytz.timezone(str(tz_))
+        else:
+            raise TypeError(
+                f"invalid tz specification: {tz_}, must be an IANA time zone "
+                "string, a non-fractional int/float UTC offset, or a "
+                "datetime.tzinfo object (including subclasses)"
+            )
 
     @classmethod
     def from_tmy(cls, tmy_metadata, tmy_data=None, **kwargs):

pvlib/tests/test_location.py

@@ -1,5 +1,6 @@
 import datetime
 from unittest.mock import ANY
+import zoneinfo
 
 import numpy as np
 from numpy import nan
@@ -27,22 +28,68 @@ def test_location_all():
     Location(32.2, -111, 'US/Arizona', 700, 'Tucson')
 
 
-@pytest.mark.parametrize('tz', [
-    pytz.timezone('US/Arizona'), 'America/Phoenix',  -7, -7.0,
-    datetime.timezone.utc
-])
-def test_location_tz(tz):
-    Location(32.2, -111, tz)
-
-
-def test_location_invalid_tz():
+@pytest.mark.parametrize(
+    'tz,tz_expected', [
+        pytest.param('UTC', 'UTC'),
+        pytest.param('Etc/GMT+5', 'Etc/GMT+5'),
+        pytest.param('US/Mountain', 'US/Mountain'),
+        pytest.param('America/Phoenix', 'America/Phoenix'),
+        pytest.param('Asia/Kathmandu', 'Asia/Kathmandu'),
+        pytest.param('Asia/Yangon', 'Asia/Yangon'),
+        pytest.param(datetime.timezone.utc, 'UTC'),
+        pytest.param(zoneinfo.ZoneInfo('Etc/GMT-7'), 'Etc/GMT-7'),
+        pytest.param(pytz.timezone('US/Arizona'), 'US/Arizona'),
+        pytest.param(-6, 'Etc/GMT+6'),
+        pytest.param(-11.0, 'Etc/GMT+11'),
+        pytest.param(12, 'Etc/GMT-12'),
+    ],
+)
+def test_location_tz(tz, tz_expected):
+    loc = Location(32.2, -111, tz)
+    assert isinstance(loc.pytz, datetime.tzinfo)  # Abstract base class.
+    assert isinstance(loc.pytz, pytz.tzinfo.BaseTzInfo)
+    assert type(loc.tz) is str
+    assert loc.tz == tz_expected
+
+
+def test_location_tz_update():
+    loc = Location(32.2, -111, -11)
+    assert loc.tz == 'Etc/GMT+11'
+    assert loc.pytz == pytz.timezone('Etc/GMT+11')
+
+    # Updating tz updates pytz.
+    loc.tz = 7
+    assert loc.tz == 'Etc/GMT-7'
+    assert loc.pytz == pytz.timezone('Etc/GMT-7')
+
+    # Updating pytz updates tz.
+    loc.pytz = pytz.timezone('US/Arizona')
+    assert loc.tz == 'US/Arizona'
+    assert loc.pytz == pytz.timezone('US/Arizona')
+
+
+@pytest.mark.parametrize(
+    'tz', [
+        'invalid',
+        'Etc/GMT+20',  # offset too large.
+        20,  # offset too large.
+    ]
+)
+def test_location_invalid_tz(tz):
     with pytest.raises(UnknownTimeZoneError):
-        Location(32.2, -111, 'invalid')
+        Location(32.2, -111, tz)
 
 
-def test_location_invalid_tz_type():
+@pytest.mark.parametrize(
+    'tz', [
+        -9.5,  # float with non-zero fractional part.
+        b"bytes not str",
+        [5],
+    ]
+)
+def test_location_invalid_tz_type(tz):
     with pytest.raises(TypeError):
-        Location(32.2, -111, [5])
+        Location(32.2, -111, tz)
 
 
 def test_location_print_all():

pvlib/tests/test_tools.py

@@ -1,9 +1,12 @@
-import pytest
+from datetime import datetime
+from zoneinfo import ZoneInfo
 
-from pvlib import tools
 import numpy as np
-import pandas as pd
 from numpy.testing import assert_allclose
+import pandas as pd
+import pytest
+
+from pvlib import location, tools
 
 
 @pytest.mark.parametrize('keys, input_dict, expected', [
@@ -144,3 +147,108 @@ def test_get_pandas_index(args, args_idx):
 def test_normalize_max2one(data_in, expected):
     result = tools.normalize_max2one(data_in)
     assert_allclose(result, expected)
+
+
+def test_localize_to_utc():
+    lat, lon = 43.2, -77.6
+    tz = "Etc/GMT+5"
+    loc = location.Location(lat, lon, tz=tz)
+    year, month, day, hour, minute, second = 1974, 6, 22, 18, 30, 15
+    hour_utc = hour + 5
+
+    # Test all combinations of supported inputs.
+    dt_time_aware_utc = datetime(
+        year, month, day, hour_utc, minute, second, tzinfo=ZoneInfo("UTC")
+    )
+    dt_time_aware = datetime(
+        year, month, day, hour, minute, second, tzinfo=ZoneInfo(tz)
+    )
+    assert tools.localize_to_utc(dt_time_aware, None) == dt_time_aware_utc
+    dt_time_naive = datetime(year, month, day, hour, minute, second)
+    assert tools.localize_to_utc(dt_time_naive, loc) == dt_time_aware_utc
+
+    # FIXME Derive timestamp strings from above variables.
+    dt_index_aware_utc = pd.DatetimeIndex(
+        [dt_time_aware_utc.strftime("%Y-%m-%dT%H:%M:%S")], tz=ZoneInfo("UTC")
+    )
+    dt_index_aware = pd.DatetimeIndex(
+        [dt_time_aware.strftime("%Y-%m-%dT%H:%M:%S")], tz=ZoneInfo(tz)
+    )
+    assert tools.localize_to_utc(dt_index_aware, None) == dt_index_aware_utc
+    dt_index_naive = pd.DatetimeIndex(
+        [dt_time_naive.strftime("%Y-%m-%dT%H:%M:%S")]
+    )
+    assert tools.localize_to_utc(dt_index_naive, loc) == dt_index_aware_utc
+
+    # Older pandas versions have wonky dtype equality check on timestamp
+    # index, so check the values as numpy.ndarray and indices one by one.
+    series_time_aware_utc_expected = pd.Series([24.42], dt_index_aware_utc)
+    series_time_aware = pd.Series([24.42], index=dt_index_aware)
+    series_time_aware_utc_got = tools.localize_to_utc(series_time_aware, None)
+    np.testing.assert_array_equal(
+        series_time_aware_utc_got.to_numpy(),
+        series_time_aware_utc_expected.to_numpy(),
+    )
+
+    for index_got, index_expected in zip(
+        series_time_aware_utc_got.index, series_time_aware_utc_expected.index
+    ):
+        assert index_got == index_expected
+
+    series_time_naive = pd.Series([24.42], index=dt_index_naive)
+    series_time_naive_utc_got = tools.localize_to_utc(series_time_naive, loc)
+    np.testing.assert_array_equal(
+        series_time_naive_utc_got.to_numpy(),
+        series_time_aware_utc_expected.to_numpy(),
+    )
+
+    for index_got, index_expected in zip(
+        series_time_naive_utc_got.index, series_time_aware_utc_expected.index
+    ):
+        assert index_got == index_expected
+
+    # Older pandas versions have wonky dtype equality check on timestamp
+    # index, so check the values as numpy.ndarray and indices one by one.
+    df_time_aware_utc_expected = pd.DataFrame([[24.42]], dt_index_aware)
+    df_time_naive = pd.DataFrame([[24.42]], index=dt_index_naive)
+    df_time_naive_utc_got = tools.localize_to_utc(df_time_naive, loc)
+    np.testing.assert_array_equal(
+        df_time_naive_utc_got.to_numpy(),
+        df_time_aware_utc_expected.to_numpy(),
+    )
+
+    for index_got, index_expected in zip(
+        df_time_naive_utc_got.index, df_time_aware_utc_expected.index
+    ):
+        assert index_got == index_expected
+
+    df_time_aware = pd.DataFrame([[24.42]], index=dt_index_aware)
+    df_time_aware_utc_got = tools.localize_to_utc(df_time_aware, None)
+    np.testing.assert_array_equal(
+        df_time_aware_utc_got.to_numpy(),
+        df_time_aware_utc_expected.to_numpy(),
+    )
+
+    for index_got, index_expected in zip(
+        df_time_aware_utc_got.index, df_time_aware_utc_expected.index
+    ):
+        assert index_got == index_expected
+
+
+def test_datetime_to_djd():
+    expected = 27201.47934027778
+    dt_aware = datetime(1974, 6, 22, 18, 30, 15, tzinfo=ZoneInfo("Etc/GMT+5"))
+    assert tools.datetime_to_djd(dt_aware) == expected
+    dt_naive_utc = datetime(1974, 6, 22, 23, 30, 15)
+    assert tools.datetime_to_djd(dt_naive_utc) == expected
+
+
+def test_djd_to_datetime():
+    djd = 27201.47934027778
+    tz = "Etc/GMT+5"
+
+    expected = datetime(1974, 6, 22, 18, 30, 15, tzinfo=ZoneInfo(tz))
+    assert tools.djd_to_datetime(djd, tz) == expected
+
+    expected = datetime(1974, 6, 22, 23, 30, 15, tzinfo=ZoneInfo("UTC"))
+    assert tools.djd_to_datetime(djd) == expected

pvlib/tools.py

@@ -119,21 +119,21 @@ def atand(number):
 
 def localize_to_utc(time, location):
     """
-    Converts or localizes a time series to UTC.
+    Converts time to UTC, localizing if necessary using location.
 
     Parameters
     ----------
     time : datetime.datetime, pandas.DatetimeIndex,
            or pandas.Series/DataFrame with a DatetimeIndex.
-    location : pvlib.Location object
+    location : pvlib.Location object (unused if time is localized)
 
     Returns
     -------
-    pandas object localized to UTC.
+    datetime.datetime or pandas object localized to UTC.
     """
     if isinstance(time, dt.datetime):
         if time.tzinfo is None:
-            time = pytz.timezone(location.tz).localize(time)
+            time = location.pytz.localize(time)
         time_utc = time.astimezone(pytz.utc)
     else:
         try: