def test_get_duration(n_timepoints, index_type):
index = _make_index(n_timepoints, index_type)
duration = _get_duration(index)
# check output type is duration type
assert isinstance(
duration,
(pd.Timedelta, pd.tseries.offsets.BaseOffset, int, np.integer))
# check integer output
duration = _get_duration(index, coerce_to_int=True)
assert isinstance(duration, (int, np.integer))
assert duration == n_timepoints - 1
def _align_seasonal(self, y):
"""Align seasonal components with y's time index"""
shift = (-_get_duration(
y.index[0],
self._y_index[0],
coerce_to_int=True,
unit=_get_freq(self._y_index),
) % self.sp)
return np.resize(np.roll(self.seasonal_, shift=shift), y.shape[0])
def test_get_duration(n_timepoints, index_type):
"""Test getting of duration."""
if index_type != "timedelta":
index = _make_index(n_timepoints, index_type)
duration = _get_duration(index)
# check output type is duration type
assert isinstance(
duration,
(pd.Timedelta, pd.tseries.offsets.BaseOffset, int, np.integer))
# check integer output
duration = _get_duration(index, coerce_to_int=True)
assert isinstance(duration, (int, np.integer))
assert duration == n_timepoints - 1
else:
match = "index_class: timedelta is not supported"
with pytest.raises(ValueError, match=match):
_make_index(n_timepoints, index_type)
def fit(self, y_train, fh=None, X_train=None):
"""Fit to training data.
Parameters
----------
y_train : pd.Series
Target time series with which to fit the forecaster.
fh : int, list or np.array, optional (default=None)
The forecast horizon with the steps ahead to predict.
X_train : pd.DataFrame, optional (default=None)
Exogenous variables are ignored
Returns
-------
self : returns an instance of self.
"""
if X_train is not None:
raise NotImplementedError("Exogeneous variables are not "
"yet supported")
self._set_y_X(y_train, X_train)
self._set_fh(fh)
# for default regressor, set fit_intercept=False as we generate a
# dummy variable in polynomial features
if self.regressor is None:
regressor = LinearRegression(fit_intercept=False)
else:
regressor = self.regressor
# make pipeline with polynomial features
self.regressor_ = make_pipeline(
PolynomialFeatures(degree=self.degree,
include_bias=self.with_intercept),
regressor,
)
# transform data
n_timepoints = _get_duration(self._y.index, coerce_to_int=True) + 1
X_train = np.arange(n_timepoints).reshape(-1, 1)
# fit regressor
self.regressor_.fit(X_train, y_train)
self._is_fitted = True
return self
def _fit(self, y, X=None, fh=None):
"""Fit to training data.
Parameters
----------
y : pd.Series
Target time series with which to fit the forecaster.
X : pd.DataFrame, default=None
Exogenous variables are ignored
fh : int, list or np.array, default=None
The forecasters horizon with the steps ahead to to predict.
Returns
-------
self : returns an instance of self.
"""
# for default regressor, set fit_intercept=False as we generate a
# dummy variable in polynomial features
if self.regressor is None:
regressor = LinearRegression(fit_intercept=False)
else:
regressor = self.regressor
# make pipeline with polynomial features
self.regressor_ = make_pipeline(
PolynomialFeatures(degree=self.degree,
include_bias=self.with_intercept),
regressor,
)
# transform data
n_timepoints = _get_duration(self._y.index, coerce_to_int=True) + 1
X = np.arange(n_timepoints).reshape(-1, 1)
# fit regressor
self.regressor_.fit(X, y)
return self
