class TimestampOps(object):
params = [None, 'US/Eastern', pytz.UTC,
dateutil.tz.tzutc()]
param_names = ['tz']
def setup(self, tz):
self.ts = Timestamp('2017-08-25 08:16:14', tz=tz)
def time_replace_tz(self, tz):
self.ts.replace(tzinfo=pytz.timezone('US/Eastern'))
def time_replace_None(self, tz):
self.ts.replace(tzinfo=None)
def time_to_pydatetime(self, tz):
self.ts.to_pydatetime()
def time_normalize(self, tz):
self.ts.normalize()
def time_tz_convert(self, tz):
if self.ts.tz is not None:
self.ts.tz_convert(tz)
def time_tz_localize(self, tz):
if self.ts.tz is None:
self.ts.tz_localize(tz)
def test_timestamp_tz_localize_nonexistent_raise(self, tz):
# GH 8917
ts = Timestamp('2015-03-29 02:20:00')
with pytest.raises(pytz.NonExistentTimeError):
ts.tz_localize(tz, nonexistent='raise')
with pytest.raises(ValueError):
ts.tz_localize(tz, nonexistent='foo')
def test_tz_localize_errors_invalid_arg(self):
# GH 22644
tz = 'Europe/Warsaw'
ts = Timestamp('2015-03-29 02:00:00')
with pytest.raises(ValueError):
with tm.assert_produces_warning(FutureWarning):
ts.tz_localize(tz, errors='foo')
def test_tz_localize_nonexistent(self, stamp, tz):
# GH#13057
ts = Timestamp(stamp)
with pytest.raises(NonExistentTimeError):
ts.tz_localize(tz)
with pytest.raises(NonExistentTimeError):
ts.tz_localize(tz, errors='raise')
assert ts.tz_localize(tz, errors='coerce') is NaT
def test_timestamp_tz_localize_nonexistent_shift_invalid(self, offset,
tz_type):
# GH 8917, 24466
tz = tz_type + 'Europe/Warsaw'
ts = Timestamp('2015-03-29 02:20:00')
msg = "The provided timedelta will relocalize on a nonexistent time"
with pytest.raises(ValueError, match=msg):
ts.tz_localize(tz, nonexistent=timedelta(seconds=offset))
def test_tz_localize_errors_coerce(self):
# GH 22644
# make sure errors='coerce' gets mapped correctly to nonexistent
tz = 'Europe/Warsaw'
ts = Timestamp('2015-03-29 02:00:00')
with tm.assert_produces_warning(FutureWarning):
result = ts.tz_localize(tz, errors='coerce')
expected = ts.tz_localize(tz, nonexistent='NaT')
assert result is expected
def test_tz_localize_nonexistent(self, stamp, tz):
# GH#13057
ts = Timestamp(stamp)
with pytest.raises(NonExistentTimeError):
ts.tz_localize(tz)
# GH 22644
with pytest.raises(NonExistentTimeError):
with tm.assert_produces_warning(FutureWarning):
ts.tz_localize(tz, errors='raise')
with tm.assert_produces_warning(FutureWarning):
assert ts.tz_localize(tz, errors='coerce') is NaT
def test_tz_localize_ambiguous_bool(self):
# make sure that we are correctly accepting bool values as ambiguous
# GH#14402
ts = Timestamp('2015-11-01 01:00:03')
expected0 = Timestamp('2015-11-01 01:00:03-0500', tz='US/Central')
expected1 = Timestamp('2015-11-01 01:00:03-0600', tz='US/Central')
with pytest.raises(pytz.AmbiguousTimeError):
ts.tz_localize('US/Central')
result = ts.tz_localize('US/Central', ambiguous=True)
assert result == expected0
result = ts.tz_localize('US/Central', ambiguous=False)
assert result == expected1
def test_timestamp_tz_localize_explicit(self):
stamp = Timestamp("3/11/2012 04:00")
result = stamp.tz_localize(self.tz("US/Eastern"))
expected = Timestamp("3/11/2012 04:00", tz=self.tz("US/Eastern"))
self.assertEqual(result.hour, expected.hour)
self.assertEqual(result, expected)
def test_timestamp_tz_localize(self, tz):
stamp = Timestamp('3/11/2012 04:00')
result = stamp.tz_localize(tz)
expected = Timestamp('3/11/2012 04:00', tz=tz)
assert result.hour == expected.hour
assert result == expected
def test_timestamp_tz_localize(self):
stamp = Timestamp('3/11/2012 04:00')
result = stamp.tz_localize('US/Eastern')
expected = Timestamp('3/11/2012 04:00', tz='US/Eastern')
self.assertEquals(result.hour, expected.hour)
self.assertEquals(result, expected)
def test_timestamp_constructor_near_dst_boundary(self):
# GH#11481 & GH#15777
# Naive string timestamps were being localized incorrectly
# with tz_convert_single instead of tz_localize_to_utc
for tz in ['Europe/Brussels', 'Europe/Prague']:
result = Timestamp('2015-10-25 01:00', tz=tz)
expected = Timestamp('2015-10-25 01:00').tz_localize(tz)
assert result == expected
with pytest.raises(pytz.AmbiguousTimeError):
Timestamp('2015-10-25 02:00', tz=tz)
result = Timestamp('2017-03-26 01:00', tz='Europe/Paris')
expected = Timestamp('2017-03-26 01:00').tz_localize('Europe/Paris')
assert result == expected
with pytest.raises(pytz.NonExistentTimeError):
Timestamp('2017-03-26 02:00', tz='Europe/Paris')
# GH#11708
naive = Timestamp('2015-11-18 10:00:00')
result = naive.tz_localize('UTC').tz_convert('Asia/Kolkata')
expected = Timestamp('2015-11-18 15:30:00+0530', tz='Asia/Kolkata')
assert result == expected
# GH#15823
result = Timestamp('2017-03-26 00:00', tz='Europe/Paris')
expected = Timestamp('2017-03-26 00:00:00+0100', tz='Europe/Paris')
assert result == expected
result = Timestamp('2017-03-26 01:00', tz='Europe/Paris')
expected = Timestamp('2017-03-26 01:00:00+0100', tz='Europe/Paris')
assert result == expected
with pytest.raises(pytz.NonExistentTimeError):
Timestamp('2017-03-26 02:00', tz='Europe/Paris')
result = Timestamp('2017-03-26 02:00:00+0100', tz='Europe/Paris')
naive = Timestamp(result.value)
expected = naive.tz_localize('UTC').tz_convert('Europe/Paris')
assert result == expected
result = Timestamp('2017-03-26 03:00', tz='Europe/Paris')
expected = Timestamp('2017-03-26 03:00:00+0200', tz='Europe/Paris')
assert result == expected
def test_tz(self):
tstr = '2014-02-01 09:00'
ts = Timestamp(tstr)
local = ts.tz_localize('Asia/Tokyo')
assert local.hour == 9
assert local == Timestamp(tstr, tz='Asia/Tokyo')
conv = local.tz_convert('US/Eastern')
assert conv == Timestamp('2014-01-31 19:00', tz='US/Eastern')
assert conv.hour == 19
# preserves nanosecond
ts = Timestamp(tstr) + offsets.Nano(5)
local = ts.tz_localize('Asia/Tokyo')
assert local.hour == 9
assert local.nanosecond == 5
conv = local.tz_convert('US/Eastern')
assert conv.nanosecond == 5
assert conv.hour == 19
def test_tz(self):
tstr = "2014-02-01 09:00"
ts = Timestamp(tstr)
local = ts.tz_localize("Asia/Tokyo")
assert local.hour == 9
assert local == Timestamp(tstr, tz="Asia/Tokyo")
conv = local.tz_convert("US/Eastern")
assert conv == Timestamp("2014-01-31 19:00", tz="US/Eastern")
assert conv.hour == 19
# preserves nanosecond
ts = Timestamp(tstr) + offsets.Nano(5)
local = ts.tz_localize("Asia/Tokyo")
assert local.hour == 9
assert local.nanosecond == 5
conv = local.tz_convert("US/Eastern")
assert conv.nanosecond == 5
assert conv.hour == 19
def pandas_timestamp_as_eastern(dt: pd.Timestamp) -> pd.Timestamp:
""" convert a pandas date as a tz-aware datetime
"""
if dt == None: return None
if type(dt) != pd.Timestamp:
raise Exception(f"type ({type(dt)}) is not pandas.TimeStamp")
if dt.tzname() != None:
raise Exception(f"value ({dt}) is not a naive timestamp")
return dt.tz_localize(eastern_tz)
def test_tz(self):
tstr = '2014-02-01 09:00'
ts = Timestamp(tstr)
local = ts.tz_localize('Asia/Tokyo')
assert local.hour == 9
assert local == Timestamp(tstr, tz='Asia/Tokyo')
conv = local.tz_convert('US/Eastern')
assert conv == Timestamp('2014-01-31 19:00', tz='US/Eastern')
assert conv.hour == 19
# preserves nanosecond
ts = Timestamp(tstr) + offsets.Nano(5)
local = ts.tz_localize('Asia/Tokyo')
assert local.hour == 9
assert local.nanosecond == 5
conv = local.tz_convert('US/Eastern')
assert conv.nanosecond == 5
assert conv.hour == 19
def test_tz_localize_ambiguous(self):
ts = Timestamp('2014-11-02 01:00')
ts_dst = ts.tz_localize('US/Eastern', ambiguous=True)
ts_no_dst = ts.tz_localize('US/Eastern', ambiguous=False)
assert (ts_no_dst.value - ts_dst.value) / 1e9 == 3600
with pytest.raises(ValueError):
ts.tz_localize('US/Eastern', ambiguous='infer')
# GH#8025
msg = ('Cannot localize tz-aware Timestamp, '
'use tz_convert for conversions')
with pytest.raises(TypeError, match=msg):
Timestamp('2011-01-01', tz='US/Eastern').tz_localize('Asia/Tokyo')
msg = ('Cannot convert tz-naive Timestamp, '
'use tz_localize to localize')
with pytest.raises(TypeError, match=msg):
Timestamp('2011-01-01').tz_convert('Asia/Tokyo')
def test_timestamp_tz_localize_nonexistent_shift(self, start_ts, tz,
end_ts, shift, tz_type):
# GH 8917, 24466
tz = tz_type + tz
if isinstance(shift, str):
shift = "shift_" + shift
ts = Timestamp(start_ts)
result = ts.tz_localize(tz, nonexistent=shift)
expected = Timestamp(end_ts).tz_localize(tz)
assert result == expected
def test_tz_localize_ambiguous(self):
ts = Timestamp('2014-11-02 01:00')
ts_dst = ts.tz_localize('US/Eastern', ambiguous=True)
ts_no_dst = ts.tz_localize('US/Eastern', ambiguous=False)
assert (ts_no_dst.value - ts_dst.value) / 1e9 == 3600
with pytest.raises(ValueError):
ts.tz_localize('US/Eastern', ambiguous='infer')
# GH#8025
with tm.assert_raises_regex(TypeError,
'Cannot localize tz-aware Timestamp, '
'use tz_convert for conversions'):
Timestamp('2011-01-01', tz='US/Eastern').tz_localize('Asia/Tokyo')
with tm.assert_raises_regex(TypeError,
'Cannot convert tz-naive Timestamp, '
'use tz_localize to localize'):
Timestamp('2011-01-01').tz_convert('Asia/Tokyo')
def test_timestamp_tz_localize_nonexistent_shift(self, start_ts, tz,
end_ts, shift,
tz_type):
# GH 8917, 24466
tz = tz_type + tz
if isinstance(shift, str):
shift = 'shift_' + shift
ts = Timestamp(start_ts)
result = ts.tz_localize(tz, nonexistent=shift)
expected = Timestamp(end_ts).tz_localize(tz)
assert result == expected
def test_tz_localize_roundtrip(self, stamp, tz):
ts = Timestamp(stamp)
localized = ts.tz_localize(tz)
assert localized == Timestamp(stamp, tz=tz)
with pytest.raises(TypeError):
localized.tz_localize(tz)
reset = localized.tz_localize(None)
assert reset == ts
assert reset.tzinfo is None
def test_tz_localize_roundtrip(self, stamp, tz):
ts = Timestamp(stamp)
localized = ts.tz_localize(tz)
assert localized == Timestamp(stamp, tz=tz)
with pytest.raises(TypeError):
localized.tz_localize(tz)
reset = localized.tz_localize(None)
assert reset == ts
assert reset.tzinfo is None
def test_describe_with_tz(self, tz_naive_fixture):
# GH 21332
tz = tz_naive_fixture
name = str(tz_naive_fixture)
start = Timestamp(2018, 1, 1)
end = Timestamp(2018, 1, 5)
s = Series(date_range(start, end, tz=tz), name=name)
result = s.describe()
expected = Series(
[
5,
5,
s.value_counts().index[0],
1,
start.tz_localize(tz),
end.tz_localize(tz),
],
name=name,
index=["count", "unique", "top", "freq", "first", "last"],
)
tm.assert_series_equal(result, expected)
def test_describe_tz_values2(self):
tz = "CET"
s1 = Series(range(5))
start = Timestamp(2018, 1, 1)
end = Timestamp(2018, 1, 5)
s2 = Series(date_range(start, end, tz=tz))
df = DataFrame({"s1": s1, "s2": s2})
s1_ = s1.describe()
s2_ = Series(
[
5,
5,
s2.value_counts().index[0],
1,
start.tz_localize(tz),
end.tz_localize(tz),
],
index=["count", "unique", "top", "freq", "first", "last"],
)
idx = [
"count",
"unique",
"top",
"freq",
"first",
"last",
"mean",
"std",
"min",
"25%",
"50%",
"75%",
"max",
]
expected = pd.concat([s1_, s2_], axis=1, keys=["s1", "s2"]).loc[idx]
with tm.assert_produces_warning(FutureWarning):
result = df.describe(include="all")
tm.assert_frame_equal(result, expected)
def test_describe_with_tz_warns(self):
name = tz = "CET"
start = Timestamp(2018, 1, 1)
end = Timestamp(2018, 1, 5)
s = Series(date_range(start, end, tz=tz), name=name)
with tm.assert_produces_warning(FutureWarning):
result = s.describe()
expected = Series(
[
5,
5,
s.value_counts().index[0],
1,
start.tz_localize(tz),
end.tz_localize(tz),
],
name=name,
index=["count", "unique", "top", "freq", "first", "last"],
)
tm.assert_series_equal(result, expected)
def test_describe_with_tz(self, tz_naive_fixture):
# GH 21332
tz = tz_naive_fixture
name = str(tz_naive_fixture)
start = Timestamp(2018, 1, 1)
end = Timestamp(2018, 1, 5)
s = Series(date_range(start, end, tz=tz), name=name)
result = s.describe(datetime_is_numeric=True)
expected = Series(
[
5,
Timestamp(2018, 1, 3).tz_localize(tz),
start.tz_localize(tz),
s[1],
s[2],
s[3],
end.tz_localize(tz),
],
name=name,
index=["count", "mean", "min", "25%", "50%", "75%", "max"],
)
tm.assert_series_equal(result, expected)
def test_tz_localize_roundtrip(self, stamp, tz_aware_fixture):
tz = tz_aware_fixture
ts = Timestamp(stamp)
localized = ts.tz_localize(tz)
assert localized == Timestamp(stamp, tz=tz)
msg = "Cannot localize tz-aware Timestamp"
with pytest.raises(TypeError, match=msg):
localized.tz_localize(tz)
reset = localized.tz_localize(None)
assert reset == ts
assert reset.tzinfo is None
def test_tz_localize_ambiguous_compat(self):
# validate that pytz and dateutil are compat for dst
# when the transition happens
naive = Timestamp('2013-10-27 01:00:00')
pytz_zone = 'Europe/London'
dateutil_zone = 'dateutil/Europe/London'
result_pytz = naive.tz_localize(pytz_zone, ambiguous=0)
result_dateutil = naive.tz_localize(dateutil_zone, ambiguous=0)
assert result_pytz.value == result_dateutil.value
assert result_pytz.value == 1382835600000000000
if LooseVersion(dateutil.__version__) < LooseVersion('2.6.0'):
# dateutil 2.6 buggy w.r.t. ambiguous=0
# see gh-14621
# see https://github.com/dateutil/dateutil/issues/321
assert (result_pytz.to_pydatetime().tzname() ==
result_dateutil.to_pydatetime().tzname())
assert str(result_pytz) == str(result_dateutil)
elif LooseVersion(dateutil.__version__) > LooseVersion('2.6.0'):
# fixed ambiguous behavior
assert result_pytz.to_pydatetime().tzname() == 'GMT'
assert result_dateutil.to_pydatetime().tzname() == 'BST'
assert str(result_pytz) != str(result_dateutil)
# 1 hour difference
result_pytz = naive.tz_localize(pytz_zone, ambiguous=1)
result_dateutil = naive.tz_localize(dateutil_zone, ambiguous=1)
assert result_pytz.value == result_dateutil.value
assert result_pytz.value == 1382832000000000000
# dateutil < 2.6 is buggy w.r.t. ambiguous timezones
if LooseVersion(dateutil.__version__) > LooseVersion('2.5.3'):
# see gh-14621
assert str(result_pytz) == str(result_dateutil)
assert (result_pytz.to_pydatetime().tzname() ==
result_dateutil.to_pydatetime().tzname())
def test_tz_localize_ambiguous_compat(self):
# validate that pytz and dateutil are compat for dst
# when the transition happens
naive = Timestamp('2013-10-27 01:00:00')
pytz_zone = 'Europe/London'
dateutil_zone = 'dateutil/Europe/London'
result_pytz = naive.tz_localize(pytz_zone, ambiguous=0)
result_dateutil = naive.tz_localize(dateutil_zone, ambiguous=0)
assert result_pytz.value == result_dateutil.value
assert result_pytz.value == 1382835600000000000
if LooseVersion(dateutil.__version__) < LooseVersion('2.6.0'):
# dateutil 2.6 buggy w.r.t. ambiguous=0
# see gh-14621
# see https://github.com/dateutil/dateutil/issues/321
assert (result_pytz.to_pydatetime().tzname() ==
result_dateutil.to_pydatetime().tzname())
assert str(result_pytz) == str(result_dateutil)
elif LooseVersion(dateutil.__version__) > LooseVersion('2.6.0'):
# fixed ambiguous behavior
assert result_pytz.to_pydatetime().tzname() == 'GMT'
assert result_dateutil.to_pydatetime().tzname() == 'BST'
assert str(result_pytz) != str(result_dateutil)
# 1 hour difference
result_pytz = naive.tz_localize(pytz_zone, ambiguous=1)
result_dateutil = naive.tz_localize(dateutil_zone, ambiguous=1)
assert result_pytz.value == result_dateutil.value
assert result_pytz.value == 1382832000000000000
# dateutil < 2.6 is buggy w.r.t. ambiguous timezones
if LooseVersion(dateutil.__version__) > LooseVersion('2.5.3'):
# see gh-14621
assert str(result_pytz) == str(result_dateutil)
assert (result_pytz.to_pydatetime().tzname() ==
result_dateutil.to_pydatetime().tzname())
def test_constructor_coverage(self):
# float value for periods
expected = pd.interval_range(start=0, periods=10)
result = pd.interval_range(start=0, periods=10.5)
tm.assert_index_equal(result, expected)
# equivalent timestamp-like start/end
start, end = Timestamp('2017-01-01'), Timestamp('2017-01-15')
expected = pd.interval_range(start=start, end=end)
result = pd.interval_range(start=start.to_pydatetime(),
end=end.to_pydatetime())
tm.assert_index_equal(result, expected)
result = pd.interval_range(start=start.tz_localize('UTC'),
end=end.tz_localize('UTC'))
tm.assert_index_equal(result, expected)
result = pd.interval_range(start=start.asm8, end=end.asm8)
tm.assert_index_equal(result, expected)
# equivalent freq with timestamp
equiv_freq = [
'D',
Day(),
Timedelta(days=1),
timedelta(days=1),
DateOffset(days=1)
]
for freq in equiv_freq:
result = pd.interval_range(start=start, end=end, freq=freq)
tm.assert_index_equal(result, expected)
# equivalent timedelta-like start/end
start, end = Timedelta(days=1), Timedelta(days=10)
expected = pd.interval_range(start=start, end=end)
result = pd.interval_range(start=start.to_pytimedelta(),
end=end.to_pytimedelta())
tm.assert_index_equal(result, expected)
result = pd.interval_range(start=start.asm8, end=end.asm8)
tm.assert_index_equal(result, expected)
# equivalent freq with timedelta
equiv_freq = ['D', Day(), Timedelta(days=1), timedelta(days=1)]
for freq in equiv_freq:
result = pd.interval_range(start=start, end=end, freq=freq)
tm.assert_index_equal(result, expected)
def test_today(self):
ts_from_string = Timestamp('today')
ts_from_method = Timestamp.today()
ts_datetime = datetime.today()
ts_from_string_tz = Timestamp('today', tz='US/Eastern')
ts_from_method_tz = Timestamp.today(tz='US/Eastern')
# Check that the delta between the times is less than 1s (arbitrarily
# small)
delta = Timedelta(seconds=1)
assert abs(ts_from_method - ts_from_string) < delta
assert abs(ts_datetime - ts_from_method) < delta
assert abs(ts_from_method_tz - ts_from_string_tz) < delta
assert (abs(ts_from_string_tz.tz_localize(None) -
ts_from_method_tz.tz_localize(None)) < delta)
def test_today(self):
ts_from_string = Timestamp('today')
ts_from_method = Timestamp.today()
ts_datetime = datetime.today()
ts_from_string_tz = Timestamp('today', tz='US/Eastern')
ts_from_method_tz = Timestamp.today(tz='US/Eastern')
# Check that the delta between the times is less than 1s (arbitrarily
# small)
delta = Timedelta(seconds=1)
assert abs(ts_from_method - ts_from_string) < delta
assert abs(ts_datetime - ts_from_method) < delta
assert abs(ts_from_method_tz - ts_from_string_tz) < delta
assert (abs(ts_from_string_tz.tz_localize(None) -
ts_from_method_tz.tz_localize(None)) < delta)
def timestamp_tzaware(timestamp):
""" Returns the pandas Timestamp passed as a timezone (UTC) aware
Timestamp.
Args:
timestamp (pandas.Timestamp): Timezone naive Timestamp
Returns:
pandas.Timestamp: Timezone aware
"""
from pandas import Timestamp
if not isinstance(timestamp, Timestamp):
timestamp = Timestamp(timestamp)
if timestamp.tzinfo is None:
return timestamp.tz_localize(tz="UTC")
else:
return timestamp.tz_convert(tz="UTC")
def test_now(self):
# GH#9000
ts_from_string = Timestamp("now")
ts_from_method = Timestamp.now()
ts_datetime = datetime.now()
ts_from_string_tz = Timestamp("now", tz="US/Eastern")
ts_from_method_tz = Timestamp.now(tz="US/Eastern")
# Check that the delta between the times is less than 1s (arbitrarily
# small)
delta = Timedelta(seconds=1)
assert abs(ts_from_method - ts_from_string) < delta
assert abs(ts_datetime - ts_from_method) < delta
assert abs(ts_from_method_tz - ts_from_string_tz) < delta
assert (abs(
ts_from_string_tz.tz_localize(None) -
ts_from_method_tz.tz_localize(None)) < delta)
def test_constructor_coverage(self):
# float value for periods
expected = pd.interval_range(start=0, periods=10)
result = pd.interval_range(start=0, periods=10.5)
tm.assert_index_equal(result, expected)
# equivalent timestamp-like start/end
start, end = Timestamp('2017-01-01'), Timestamp('2017-01-15')
expected = pd.interval_range(start=start, end=end)
result = pd.interval_range(start=start.to_pydatetime(),
end=end.to_pydatetime())
tm.assert_index_equal(result, expected)
result = pd.interval_range(start=start.tz_localize('UTC'),
end=end.tz_localize('UTC'))
tm.assert_index_equal(result, expected)
result = pd.interval_range(start=start.asm8, end=end.asm8)
tm.assert_index_equal(result, expected)
# equivalent freq with timestamp
equiv_freq = ['D', Day(), Timedelta(days=1), timedelta(days=1),
DateOffset(days=1)]
for freq in equiv_freq:
result = pd.interval_range(start=start, end=end, freq=freq)
tm.assert_index_equal(result, expected)
# equivalent timedelta-like start/end
start, end = Timedelta(days=1), Timedelta(days=10)
expected = pd.interval_range(start=start, end=end)
result = pd.interval_range(start=start.to_pytimedelta(),
end=end.to_pytimedelta())
tm.assert_index_equal(result, expected)
result = pd.interval_range(start=start.asm8, end=end.asm8)
tm.assert_index_equal(result, expected)
# equivalent freq with timedelta
equiv_freq = ['D', Day(), Timedelta(days=1), timedelta(days=1)]
for freq in equiv_freq:
result = pd.interval_range(start=start, end=end, freq=freq)
tm.assert_index_equal(result, expected)
def create_aligned_timestamp(time):
""" Align the passed datetime / Timestamp object to the minute
interval for use in dateranges and overlap checks.
Args:
time (str, pandas.Timestamp)
Returns:
pandas.Timestamp: Timestamp aligned to minute
with UTC timezone
"""
from pandas import Timedelta, Timestamp
if not isinstance(time, Timestamp):
time = Timestamp(ts_input=time)
if time.tzinfo is None:
t = time.tz_localize(tz="UTC")
else:
t = time.tz_convert(tz="UTC")
t -= Timedelta(f"{t.second} s")
return t
def test_tz_conversion_freq(self, tz_naive_fixture):
# GH25241
t1 = Timestamp('2019-01-01 10:00', freq='H')
assert t1.tz_localize(tz=tz_naive_fixture).freq == t1.freq
t2 = Timestamp('2019-01-02 12:00', tz='UTC', freq='T')
assert t2.tz_convert(tz='UTC').freq == t2.freq
def test_timestamp_tz_localize_nonexistent_NaT(self, tz):
# GH 8917
ts = Timestamp("2015-03-29 02:20:00")
result = ts.tz_localize(tz, nonexistent="NaT")
assert result is NaT
def test_tz_localize_errors_ambiguous(self):
# GH#13057
ts = Timestamp('2015-11-1 01:00')
with pytest.raises(AmbiguousTimeError):
ts.tz_localize('US/Pacific', errors='coerce')
def test_tz_localize_errors_ambiguous(self):
# GH#13057
ts = Timestamp('2015-11-1 01:00')
with pytest.raises(AmbiguousTimeError):
ts.tz_localize('US/Pacific', errors='coerce')
def test_tz_localize_ambiguous_raise(self):
# GH#13057
ts = Timestamp("2015-11-1 01:00")
msg = "Cannot infer dst time from 2015-11-01 01:00:00,"
with pytest.raises(AmbiguousTimeError, match=msg):
ts.tz_localize("US/Pacific", ambiguous="raise")
def test_replace_tzinfo_equiv_tz_localize_none(self):
# GH#14621, GH#7825
# assert conversion to naive is the same as replacing tzinfo with None
ts = Timestamp('2013-11-03 01:59:59.999999-0400', tz='US/Eastern')
assert ts.tz_localize(None) == ts.replace(tzinfo=None)
class Holiday:
"""
Class that defines a holiday with start/end dates and rules
for observance.
"""
def __init__(
self,
name,
year=None,
month=None,
day=None,
offset=None,
observance=None,
start_date=None,
end_date=None,
days_of_week=None,
):
"""
Parameters
----------
name : str
Name of the holiday , defaults to class name
offset : array of pandas.tseries.offsets or
class from pandas.tseries.offsets
computes offset from date
observance: function
computes when holiday is given a pandas Timestamp
days_of_week:
provide a tuple of days e.g (0,1,2,3,) for Monday Through Thursday
Monday=0,..,Sunday=6
Examples
--------
>>> from dateutil.relativedelta import MO
>>> USMemorialDay = pd.tseries.holiday.Holiday(
... "Memorial Day", month=5, day=31, offset=pd.DateOffset(weekday=MO(-1))
... )
>>> USMemorialDay
Holiday: Memorial Day (month=5, day=31, offset=<DateOffset: weekday=MO(-1)>)
>>> USLaborDay = pd.tseries.holiday.Holiday(
... "Labor Day", month=9, day=1, offset=pd.DateOffset(weekday=MO(1))
... )
>>> USLaborDay
Holiday: Labor Day (month=9, day=1, offset=<DateOffset: weekday=MO(+1)>)
>>> July3rd = pd.tseries.holiday.Holiday("July 3rd", month=7, day=3)
>>> July3rd
Holiday: July 3rd (month=7, day=3, )
>>> NewYears = pd.tseries.holiday.Holiday(
... "New Years Day", month=1, day=1,
... observance=pd.tseries.holiday.nearest_workday
... )
>>> NewYears # doctest: +SKIP
Holiday: New Years Day (
month=1, day=1, observance=<function nearest_workday at 0x66545e9bc440>
)
>>> July3rd = pd.tseries.holiday.Holiday(
... "July 3rd", month=7, day=3,
... days_of_week=(0, 1, 2, 3)
... )
>>> July3rd
Holiday: July 3rd (month=7, day=3, )
"""
if offset is not None and observance is not None:
raise NotImplementedError("Cannot use both offset and observance.")
self.name = name
self.year = year
self.month = month
self.day = day
self.offset = offset
self.start_date = (Timestamp(start_date)
if start_date is not None else start_date)
self.end_date = Timestamp(
end_date) if end_date is not None else end_date
self.observance = observance
assert days_of_week is None or type(days_of_week) == tuple
self.days_of_week = days_of_week
def __repr__(self) -> str:
info = ""
if self.year is not None:
info += f"year={self.year}, "
info += f"month={self.month}, day={self.day}, "
if self.offset is not None:
info += f"offset={self.offset}"
if self.observance is not None:
info += f"observance={self.observance}"
repr = f"Holiday: {self.name} ({info})"
return repr
def dates(self, start_date, end_date, return_name=False):
"""
Calculate holidays observed between start date and end date
Parameters
----------
start_date : starting date, datetime-like, optional
end_date : ending date, datetime-like, optional
return_name : bool, optional, default=False
If True, return a series that has dates and holiday names.
False will only return dates.
"""
start_date = Timestamp(start_date)
end_date = Timestamp(end_date)
filter_start_date = start_date
filter_end_date = end_date
if self.year is not None:
dt = Timestamp(datetime(self.year, self.month, self.day))
if return_name:
return Series(self.name, index=[dt])
else:
return [dt]
dates = self._reference_dates(start_date, end_date)
holiday_dates = self._apply_rule(dates)
if self.days_of_week is not None:
holiday_dates = holiday_dates[np.in1d(holiday_dates.dayofweek,
self.days_of_week)]
if self.start_date is not None:
filter_start_date = max(
self.start_date.tz_localize(filter_start_date.tz),
filter_start_date)
if self.end_date is not None:
filter_end_date = min(
self.end_date.tz_localize(filter_end_date.tz), filter_end_date)
holiday_dates = holiday_dates[(holiday_dates >= filter_start_date)
& (holiday_dates <= filter_end_date)]
if return_name:
return Series(self.name, index=holiday_dates)
return holiday_dates
def _reference_dates(self, start_date, end_date):
"""
Get reference dates for the holiday.
Return reference dates for the holiday also returning the year
prior to the start_date and year following the end_date. This ensures
that any offsets to be applied will yield the holidays within
the passed in dates.
"""
if self.start_date is not None:
start_date = self.start_date.tz_localize(start_date.tz)
if self.end_date is not None:
end_date = self.end_date.tz_localize(start_date.tz)
year_offset = DateOffset(years=1)
reference_start_date = Timestamp(
datetime(start_date.year - 1, self.month, self.day))
reference_end_date = Timestamp(
datetime(end_date.year + 1, self.month, self.day))
# Don't process unnecessary holidays
dates = date_range(
start=reference_start_date,
end=reference_end_date,
freq=year_offset,
tz=start_date.tz,
)
return dates
def _apply_rule(self, dates):
"""
Apply the given offset/observance to a DatetimeIndex of dates.
Parameters
----------
dates : DatetimeIndex
Dates to apply the given offset/observance rule
Returns
-------
Dates with rules applied
"""
if self.observance is not None:
return dates.map(lambda d: self.observance(d))
if self.offset is not None:
if not isinstance(self.offset, list):
offsets = [self.offset]
else:
offsets = self.offset
for offset in offsets:
# if we are adding a non-vectorized value
# ignore the PerformanceWarnings:
with warnings.catch_warnings():
warnings.simplefilter("ignore", PerformanceWarning)
dates += offset
return dates
class Holiday(object):
"""
Class that defines a holiday with start/end dates and rules
for observance.
"""
def __init__(self,
name,
year=None,
month=None,
day=None,
offset=None,
observance=None,
start_date=None,
end_date=None,
days_of_week=None):
"""
Parameters
----------
name : str
Name of the holiday , defaults to class name
offset : array of pandas.tseries.offsets or
class from pandas.tseries.offsets
computes offset from date
observance: function
computes when holiday is given a pandas Timestamp
days_of_week:
provide a tuple of days e.g (0,1,2,3,) for Monday Through Thursday
Monday=0,..,Sunday=6
Examples
--------
>>> from pandas.tseries.holiday import Holiday, nearest_workday
>>> from pandas import DateOffset
>>> from dateutil.relativedelta import MO
>>> USMemorialDay = Holiday('MemorialDay', month=5, day=24,
offset=DateOffset(weekday=MO(1)))
>>> USLaborDay = Holiday('Labor Day', month=9, day=1,
offset=DateOffset(weekday=MO(1)))
>>> July3rd = Holiday('July 3rd', month=7, day=3,)
>>> NewYears = Holiday('New Years Day', month=1, day=1,
observance=nearest_workday),
>>> July3rd = Holiday('July 3rd', month=7, day=3,
days_of_week=(0, 1, 2, 3))
"""
if offset is not None and observance is not None:
raise NotImplementedError("Cannot use both offset and observance.")
self.name = name
self.year = year
self.month = month
self.day = day
self.offset = offset
self.start_date = Timestamp(
start_date) if start_date is not None else start_date
self.end_date = Timestamp(
end_date) if end_date is not None else end_date
self.observance = observance
assert (days_of_week is None or type(days_of_week) == tuple)
self.days_of_week = days_of_week
def __repr__(self):
info = ''
if self.year is not None:
info += 'year=%s, ' % self.year
info += 'month=%s, day=%s, ' % (self.month, self.day)
if self.offset is not None:
info += 'offset=%s' % self.offset
if self.observance is not None:
info += 'observance=%s' % self.observance
repr = 'Holiday: %s (%s)' % (self.name, info)
return repr
def dates(self, start_date, end_date, return_name=False):
"""
Calculate holidays observed between start date and end date
Parameters
----------
start_date : starting date, datetime-like, optional
end_date : ending date, datetime-like, optional
return_name : bool, optional, default=False
If True, return a series that has dates and holiday names.
False will only return dates.
"""
start_date = Timestamp(start_date)
end_date = Timestamp(end_date)
filter_start_date = start_date
filter_end_date = end_date
if self.year is not None:
dt = Timestamp(datetime(self.year, self.month, self.day))
if return_name:
return Series(self.name, index=[dt])
else:
return [dt]
dates = self._reference_dates(start_date, end_date)
holiday_dates = self._apply_rule(dates)
if self.days_of_week is not None:
holiday_dates = holiday_dates[np.in1d(holiday_dates.dayofweek,
self.days_of_week)]
if self.start_date is not None:
filter_start_date = max(
self.start_date.tz_localize(filter_start_date.tz),
filter_start_date)
if self.end_date is not None:
filter_end_date = min(
self.end_date.tz_localize(filter_end_date.tz), filter_end_date)
holiday_dates = holiday_dates[(holiday_dates >= filter_start_date)
& (holiday_dates <= filter_end_date)]
if return_name:
return Series(self.name, index=holiday_dates)
return holiday_dates
def _reference_dates(self, start_date, end_date):
"""
Get reference dates for the holiday.
Return reference dates for the holiday also returning the year
prior to the start_date and year following the end_date. This ensures
that any offsets to be applied will yield the holidays within
the passed in dates.
"""
if self.start_date is not None:
start_date = self.start_date.tz_localize(start_date.tz)
if self.end_date is not None:
end_date = self.end_date.tz_localize(start_date.tz)
year_offset = DateOffset(years=1)
reference_start_date = Timestamp(
datetime(start_date.year - 1, self.month, self.day))
reference_end_date = Timestamp(
datetime(end_date.year + 1, self.month, self.day))
# Don't process unnecessary holidays
dates = DatetimeIndex(start=reference_start_date,
end=reference_end_date,
freq=year_offset,
tz=start_date.tz)
return dates
def _apply_rule(self, dates):
"""
Apply the given offset/observance to a DatetimeIndex of dates.
Parameters
----------
dates : DatetimeIndex
Dates to apply the given offset/observance rule
Returns
-------
Dates with rules applied
"""
if self.observance is not None:
return dates.map(lambda d: self.observance(d))
if self.offset is not None:
if not isinstance(self.offset, list):
offsets = [self.offset]
else:
offsets = self.offset
for offset in offsets:
# if we are adding a non-vectorized value
# ignore the PerformanceWarnings:
with warnings.catch_warnings(record=True):
dates += offset
return dates
def test_tz_conversion_freq(self, tz_naive_fixture):
# GH25241
t1 = Timestamp('2019-01-01 10:00', freq='H')
assert t1.tz_localize(tz=tz_naive_fixture).freq == t1.freq
t2 = Timestamp('2019-01-02 12:00', tz='UTC', freq='T')
assert t2.tz_convert(tz='UTC').freq == t2.freq
def test_timestamp_tz_localize_nonexistent_shift(self, tz):
# GH 8917
ts = Timestamp('2015-03-29 02:20:00')
result = ts.tz_localize(tz, nonexistent='shift')
expected = Timestamp('2015-03-29 03:00:00').tz_localize(tz)
assert result == expected
def test_tz_localize_errors_ambiguous(self):
# GH#13057
ts = Timestamp('2015-11-1 01:00')
with pytest.raises(AmbiguousTimeError):
with tm.assert_produces_warning(FutureWarning):
ts.tz_localize('US/Pacific', errors='coerce')
def test_tz_localize_ambiguous_raise(self):
# GH#13057
ts = Timestamp("2015-11-1 01:00")
with pytest.raises(AmbiguousTimeError):
ts.tz_localize("US/Pacific", ambiguous="raise")
def test_tz_conversion_freq(self, tz_naive_fixture):
# GH25241
t1 = Timestamp("2019-01-01 10:00", freq="H")
assert t1.tz_localize(tz=tz_naive_fixture).freq == t1.freq
t2 = Timestamp("2019-01-02 12:00", tz="UTC", freq="T")
assert t2.tz_convert(tz="UTC").freq == t2.freq
def test_timestamp_tz_localize_nonexistent_shift(self, tz):
# GH 8917
ts = Timestamp('2015-03-29 02:20:00')
result = ts.tz_localize(tz, nonexistent='shift')
expected = Timestamp('2015-03-29 03:00:00').tz_localize(tz)
assert result == expected
def test_tz_localize_errors_ambiguous(self):
# GH#13057
ts = Timestamp('2015-11-1 01:00')
with pytest.raises(AmbiguousTimeError):
with tm.assert_produces_warning(FutureWarning):
ts.tz_localize('US/Pacific', errors='coerce')
class Holiday(object):
"""
Class that defines a holiday with start/end dates and rules
for observance.
"""
def __init__(self, name, year=None, month=None, day=None, offset=None,
observance=None, start_date=None, end_date=None,
days_of_week=None):
"""
Parameters
----------
name : str
Name of the holiday , defaults to class name
offset : array of pandas.tseries.offsets or
class from pandas.tseries.offsets
computes offset from date
observance: function
computes when holiday is given a pandas Timestamp
days_of_week:
provide a tuple of days e.g (0,1,2,3,) for Monday Through Thursday
Monday=0,..,Sunday=6
Examples
--------
>>> from pandas.tseries.holiday import Holiday, nearest_workday
>>> from pandas import DateOffset
>>> from dateutil.relativedelta import MO
>>> USMemorialDay = Holiday('MemorialDay', month=5, day=24,
offset=DateOffset(weekday=MO(1)))
>>> USLaborDay = Holiday('Labor Day', month=9, day=1,
offset=DateOffset(weekday=MO(1)))
>>> July3rd = Holiday('July 3rd', month=7, day=3,)
>>> NewYears = Holiday('New Years Day', month=1, day=1,
observance=nearest_workday),
>>> July3rd = Holiday('July 3rd', month=7, day=3,
days_of_week=(0, 1, 2, 3))
"""
if offset is not None and observance is not None:
raise NotImplementedError("Cannot use both offset and observance.")
self.name = name
self.year = year
self.month = month
self.day = day
self.offset = offset
self.start_date = Timestamp(
start_date) if start_date is not None else start_date
self.end_date = Timestamp(
end_date) if end_date is not None else end_date
self.observance = observance
assert (days_of_week is None or type(days_of_week) == tuple)
self.days_of_week = days_of_week
def __repr__(self):
info = ''
if self.year is not None:
info += 'year={year}, '.format(year=self.year)
info += 'month={mon}, day={day}, '.format(mon=self.month, day=self.day)
if self.offset is not None:
info += 'offset={offset}'.format(offset=self.offset)
if self.observance is not None:
info += 'observance={obs}'.format(obs=self.observance)
repr = 'Holiday: {name} ({info})'.format(name=self.name, info=info)
return repr
def dates(self, start_date, end_date, return_name=False):
"""
Calculate holidays observed between start date and end date
Parameters
----------
start_date : starting date, datetime-like, optional
end_date : ending date, datetime-like, optional
return_name : bool, optional, default=False
If True, return a series that has dates and holiday names.
False will only return dates.
"""
start_date = Timestamp(start_date)
end_date = Timestamp(end_date)
filter_start_date = start_date
filter_end_date = end_date
if self.year is not None:
dt = Timestamp(datetime(self.year, self.month, self.day))
if return_name:
return Series(self.name, index=[dt])
else:
return [dt]
dates = self._reference_dates(start_date, end_date)
holiday_dates = self._apply_rule(dates)
if self.days_of_week is not None:
holiday_dates = holiday_dates[np.in1d(holiday_dates.dayofweek,
self.days_of_week)]
if self.start_date is not None:
filter_start_date = max(self.start_date.tz_localize(
filter_start_date.tz), filter_start_date)
if self.end_date is not None:
filter_end_date = min(self.end_date.tz_localize(
filter_end_date.tz), filter_end_date)
holiday_dates = holiday_dates[(holiday_dates >= filter_start_date) &
(holiday_dates <= filter_end_date)]
if return_name:
return Series(self.name, index=holiday_dates)
return holiday_dates
def _reference_dates(self, start_date, end_date):
"""
Get reference dates for the holiday.
Return reference dates for the holiday also returning the year
prior to the start_date and year following the end_date. This ensures
that any offsets to be applied will yield the holidays within
the passed in dates.
"""
if self.start_date is not None:
start_date = self.start_date.tz_localize(start_date.tz)
if self.end_date is not None:
end_date = self.end_date.tz_localize(start_date.tz)
year_offset = DateOffset(years=1)
reference_start_date = Timestamp(
datetime(start_date.year - 1, self.month, self.day))
reference_end_date = Timestamp(
datetime(end_date.year + 1, self.month, self.day))
# Don't process unnecessary holidays
dates = DatetimeIndex(start=reference_start_date,
end=reference_end_date,
freq=year_offset, tz=start_date.tz)
return dates
def _apply_rule(self, dates):
"""
Apply the given offset/observance to a DatetimeIndex of dates.
Parameters
----------
dates : DatetimeIndex
Dates to apply the given offset/observance rule
Returns
-------
Dates with rules applied
"""
if self.observance is not None:
return dates.map(lambda d: self.observance(d))
if self.offset is not None:
if not isinstance(self.offset, list):
offsets = [self.offset]
else:
offsets = self.offset
for offset in offsets:
# if we are adding a non-vectorized value
# ignore the PerformanceWarnings:
with warnings.catch_warnings(record=True):
dates += offset
return dates
def test_tz_localize_nonexistent_invalid_arg(self):
# GH 22644
tz = "Europe/Warsaw"
ts = Timestamp("2015-03-29 02:00:00")
with pytest.raises(ValueError):
ts.tz_localize(tz, nonexistent="foo")
class HolidayWithFilter(object):
"""
Class that defines a holiday with start/end dates and rules
for observance.
"""
def __init__(self,
name,
year=None,
month=None,
day=None,
offset=None,
observance=None,
start_date=None,
end_date=None,
days_of_week=None,
year_filter=None,
year_mask=None):
"""
Parameters
----------
name : str
Name of the holiday , defaults to class name
offset : array of pandas.tseries.offsets or
class from pandas.tseries.offsets
computes offset from date
observance: function
computes when holiday is given a pandas Timestamp
days_of_week:
provide a tuple of days e.g (0,1,2,3,) for Monday Through Thursday
Monday=0,..,Sunday=6
"""
if offset is not None and observance is not None:
raise NotImplementedError("Cannot use both offset and observance.")
self.name = name
self.year = year
self.month = month
self.day = day
self.offset = offset
self.start_date = Timestamp(
start_date) if start_date is not None else start_date
self.end_date = Timestamp(
end_date) if end_date is not None else end_date
self.observance = observance
assert (days_of_week is None or type(days_of_week) == tuple)
self.days_of_week = days_of_week
self.year_filter = year_filter
self.year_mask = year_mask
def __repr__(self):
info = ''
if self.year is not None:
info += 'year={year}, '.format(year=self.year)
info += 'month={mon}, day={day}, '.format(mon=self.month, day=self.day)
if self.offset is not None:
info += 'offset={offset}'.format(offset=self.offset)
if self.observance is not None:
info += 'observance={obs}'.format(obs=self.observance)
if self.year_filter is not None:
info += 'year_filter={yrf}'.format(yrf=self.year_filter)
if self.year_mask is not None:
info += 'year_mask={yrm}'.format(yrf=self.year_mask)
repr = 'Holiday: {name} ({info})'.format(name=self.name, info=info)
return repr
def dates(self, start_date, end_date, return_name=False):
"""
Calculate holidays observed between start date and end date
Parameters
----------
start_date : starting date, datetime-like, optional
end_date : ending date, datetime-like, optional
return_name : bool, optional, default=False
If True, return a series that has dates and holiday names.
False will only return dates.
"""
start_date = Timestamp(start_date)
end_date = Timestamp(end_date)
filter_start_date = start_date
filter_end_date = end_date
if self.year is not None:
dt = Timestamp(datetime(self.year, self.month, self.day))
if return_name:
return Series(self.name, index=[dt])
else:
return [dt]
dates = self._reference_dates(start_date, end_date)
holiday_dates = self._apply_rule(dates)
if self.days_of_week is not None:
holiday_dates = holiday_dates[np.in1d(holiday_dates.dayofweek,
self.days_of_week)]
if self.start_date is not None:
filter_start_date = max(
self.start_date.tz_localize(filter_start_date.tz),
filter_start_date)
if self.end_date is not None:
filter_end_date = min(
self.end_date.tz_localize(filter_end_date.tz), filter_end_date)
holiday_dates = holiday_dates[(holiday_dates >= filter_start_date)
& (holiday_dates <= filter_end_date)]
if return_name:
return Series(self.name, index=holiday_dates)
return holiday_dates
def _reference_dates(self, start_date, end_date):
"""
Get reference dates for the holiday.
Return reference dates for the holiday also returning the year
prior to the start_date and year following the end_date. This ensures
that any offsets to be applied will yield the holidays within
the passed in dates.
"""
if self.start_date is not None:
start_date = self.start_date.tz_localize(start_date.tz)
if self.end_date is not None:
end_date = self.end_date.tz_localize(start_date.tz)
year_offset = DateOffset(years=1)
reference_start_date = Timestamp(
datetime(start_date.year - 1, self.month, self.day))
reference_end_date = Timestamp(
datetime(end_date.year + 1, self.month, self.day))
# Don't process unnecessary holidays
dates = DatetimeIndex(start=reference_start_date,
end=reference_end_date,
freq=year_offset,
tz=start_date.tz)
if self.year_filter is not None:
dates = dates[~dates.year.isin(self.year_filter)]
if self.year_mask is not None:
dates = dates[dates.year.isin(self.year_mask)]
return dates
def _apply_rule(self, dates):
"""
Apply the given offset/observance to a DatetimeIndex of dates.
Parameters
----------
dates : DatetimeIndex
Dates to apply the given offset/observance rule
Returns
-------
Dates with rules applied
"""
if self.observance is not None:
return dates.map(lambda d: self.observance(d))
if self.offset is not None:
if not isinstance(self.offset, list):
offsets = [self.offset]
else:
offsets = self.offset
for offset in offsets:
# if we are adding a non-vectorized value
# ignore the PerformanceWarnings:
with warnings.catch_warnings():
warnings.simplefilter("ignore", PerformanceWarning)
dates += offset
return dates
def test_replace_tzinfo_equiv_tz_localize_none(self):
# GH#14621, GH#7825
# assert conversion to naive is the same as replacing tzinfo with None
ts = Timestamp("2013-11-03 01:59:59.999999-0400", tz="US/Eastern")
assert ts.tz_localize(None) == ts.replace(tzinfo=None)
def test_timestamp_tz_localize_nonexistent_NaT(self, tz):
# GH 8917
ts = Timestamp('2015-03-29 02:20:00')
result = ts.tz_localize(tz, nonexistent='NaT')
assert result is NaT