def seasonal_features(cls, freq: str) -> List[TimeFeature]:
offset = to_offset(freq)
if offset.name == "M":
return [MonthOfYear(normalized=False)]
elif offset.name == "W-SUN":
return [WeekOfYear(normalized=False)]
elif offset.name == "D":
return [DayOfWeek(normalized=False)]
elif offset.name == "B": # TODO: check this case
return [DayOfWeek(normalized=False)]
elif offset.name == "H":
return [HourOfDay(normalized=False), DayOfWeek(normalized=False)]
elif offset.name == "T":
return [
MinuteOfHour(normalized=False),
HourOfDay(normalized=False),
]
else:
RuntimeError(f"Unsupported frequency {offset.name}")
return []
DayOfYearIndex,
WeekOfYearIndex,
MonthOfYearIndex,
)
@pytest.mark.parametrize(
"feature, index",
[
(
MinuteOfHour(),
pd.date_range(
"01-01-2015 00:00:00", periods=60 * 2 * 24, freq="1min"),
),
(
HourOfDay(),
pd.date_range("01-01-2015 00:00:00", periods=14 * 24, freq="1h"),
),
(DayOfWeek(), pd.date_range("01-01-2015", periods=365 * 5, freq="D")),
(DayOfMonth(), pd.date_range("01-01-2015", periods=365 * 5, freq="D")),
(DayOfYear(), pd.date_range("01-01-2015", periods=365 * 5, freq="D")),
(WeekOfYear(), pd.date_range("01-01-2015", periods=53 * 5, freq="W")),
(MonthOfYear(), pd.date_range("01-01-2015", periods=12 * 5, freq="M")),
(Constant(), pd.date_range("01-01-2015", periods=5, freq="A")),
],
)
def test_feature_normalized_bounds(feature: TimeFeature,
index: pd.DatetimeIndex):
values = feature(index)
assert isinstance(values, np.ndarray)
for v in values:
TabularEstimator, )
from gluonts.model.predictor import Predictor
from gluonts.time_feature import TimeFeature, HourOfDay, DayOfWeek, MonthOfYear
@pytest.mark.parametrize(
"series, time_features, lag_indices, past_data, expected_df",
[
(
pd.Series(
list(range(5)),
index=pd.period_range(
"2020-12-31 22:00:00", freq="H", periods=5),
),
[MonthOfYear(), DayOfWeek(),
HourOfDay()],
[1, 2, 5],
None,
pd.DataFrame(
{
"MonthOfYear": [0.5, 0.5, -0.5, -0.5, -0.5],
"DayOfWeek": [
0.0,
0.0,
4.0 / 6 - 0.5,
4.0 / 6 - 0.5,
4.0 / 6 - 0.5,
],
"HourOfDay": [
22.0 / 23 - 0.5,
0.5,
)
from gluonts.model.predictor import Predictor
from gluonts.time_feature import TimeFeature, HourOfDay, DayOfWeek, MonthOfYear
@pytest.mark.parametrize(
"series, time_features, lag_indices, past_data, expected_df",
[
(
pd.Series(
list(range(5)),
index=pd.date_range(
"2020-12-31 22:00:00", freq="H", periods=5
),
),
[MonthOfYear(), DayOfWeek(), HourOfDay()],
[1, 2, 5],
None,
pd.DataFrame(
{
"MonthOfYear": [0.5, 0.5, -0.5, -0.5, -0.5],
"DayOfWeek": [
0.0,
0.0,
4.0 / 6 - 0.5,
4.0 / 6 - 0.5,
4.0 / 6 - 0.5,
],
"HourOfDay": [
22.0 / 23 - 0.5,
0.5,