def is_stationary(self):
"""(bool) Is the reduced autoregressive lag poylnomial stationary."""
validate_basic(self.ar_params,
self.k_ar_params,
title='AR coefficients')
validate_basic(self.seasonal_ar_params,
self.k_seasonal_ar_params,
title='seasonal AR coefficients')
ar_stationary = True
seasonal_ar_stationary = True
if self.k_ar_params > 0:
ar_stationary = is_invertible(self.ar_poly.coef)
if self.k_seasonal_ar_params > 0:
seasonal_ar_stationary = is_invertible(self.seasonal_ar_poly.coef)
return ar_stationary and seasonal_ar_stationary
def is_invertible(self):
"""(bool) Is the reduced moving average lag poylnomial invertible."""
# Short-circuit if there is no MA component
validate_basic(self.ma_params,
self.k_ma_params,
title='MA coefficients')
validate_basic(self.seasonal_ma_params,
self.k_seasonal_ma_params,
title='seasonal MA coefficients')
ma_stationary = True
seasonal_ma_stationary = True
if self.k_ma_params > 0:
ma_stationary = is_invertible(self.ma_poly.coef)
if self.k_seasonal_ma_params > 0:
seasonal_ma_stationary = is_invertible(self.seasonal_ma_poly.coef)
return ma_stationary and seasonal_ma_stationary
def split_params(self, params, allow_infnan=False):
"""
Split parameter array by type into dictionary.
Parameters
----------
params : array_like
Array of model parameters.
allow_infnan : bool, optional
Whether or not to allow `params` to contain -np.Inf, np.Inf, and
np.nan. Default is False.
Returns
-------
split_params : dict
Dictionary with keys 'exog_params', 'ar_params', 'ma_params',
'seasonal_ar_params', 'seasonal_ma_params', and (unless
`concentrate_scale=True`) 'sigma2'. Values are the parameters
associated with the key, based on the `params` argument.
Examples
--------
>>> spec = SARIMAXSpecification(ar_order=1)
>>> spec.split_params([0.5, 4])
{'exog_params': array([], dtype=float64),
'ar_params': array([0.5]),
'ma_params': array([], dtype=float64),
'seasonal_ar_params': array([], dtype=float64),
'seasonal_ma_params': array([], dtype=float64),
'sigma2': 4.0}
"""
params = validate_basic(params,
self.k_params,
allow_infnan=allow_infnan,
title='joint parameters')
ix = [
self.k_exog_params, self.k_ar_params, self.k_ma_params,
self.k_seasonal_ar_params, self.k_seasonal_ma_params
]
names = [
'exog_params', 'ar_params', 'ma_params', 'seasonal_ar_params',
'seasonal_ma_params'
]
if not self.concentrate_scale:
ix.append(1)
names.append('sigma2')
ix = np.cumsum(ix)
out = dict(zip(names, np.split(params, ix)))
if 'sigma2' in out:
out['sigma2'] = out['sigma2'].item()
return out
def test_validate_basic():
# Valid parameters
assert_equal(validate_basic([], 0, title='test'), [])
assert_equal(validate_basic(0, 1), [0])
assert_equal(validate_basic([0], 1), [0])
assert_equal(validate_basic(np.array([1.2, 0.5 + 1j]), 2),
np.array([1.2, 0.5 + 1j]))
assert_equal(
validate_basic([np.nan, -np.inf, np.inf], 3, allow_infnan=True),
[np.nan, -np.inf, np.inf])
# Invalid parameters
assert_raises(ValueError, validate_basic, [], 1, title='test')
assert_raises(ValueError, validate_basic, 0, 0)
assert_raises(ValueError, validate_basic, 'a', 1)
assert_raises(ValueError, validate_basic, None, 1)
assert_raises(ValueError, validate_basic, np.nan, 1)
assert_raises(ValueError, validate_basic, np.inf, 1)
assert_raises(ValueError, validate_basic, -np.inf, 1)
assert_raises(ValueError, validate_basic, [1, 2], 1)
def ma_poly(self, value):
# Convert from the polynomial to the parameters, and set that way
if isinstance(value, Polynomial):
value = value.coef
value = validate_basic(value,
self.spec.max_ma_order + 1,
title='MA polynomial')
if value[0] != 1:
raise ValueError('MA polynomial constant must be equal to 1.')
ma_params = []
for i in range(1, self.spec.max_ma_order + 1):
if i in self.spec.ma_lags:
ma_params.append(value[i])
elif value[i] != 0:
raise ValueError('MA polynomial includes non-zero values'
' for lags that are excluded in the'
' specification.')
self.ma_params = ma_params
def seasonal_ar_poly(self, value):
s = self.spec.seasonal_periods
# Note: assume that we are given coefficients from the full polynomial
# Convert from the polynomial to the parameters, and set that way
if isinstance(value, Polynomial):
value = value.coef
value = validate_basic(value,
1 + s * self.spec.max_seasonal_ar_order,
title='seasonal AR polynomial')
if value[0] != 1:
raise ValueError('Polynomial constant must be equal to 1.')
seasonal_ar_params = []
for i in range(1, self.spec.max_seasonal_ar_order + 1):
if i in self.spec.seasonal_ar_lags:
seasonal_ar_params.append(-value[s * i])
elif value[s * i] != 0:
raise ValueError('AR polynomial includes non-zero values'
' for lags that are excluded in the'
' specification.')
self.seasonal_ar_params = seasonal_ar_params
def ar_params(self, value):
if np.isscalar(value):
value = [value] * self.k_ar_params
self._params_split['ar_params'] = validate_basic(
value, self.k_ar_params, title='AR coefficients')
self._params = None
def exog_params(self, value):
if np.isscalar(value):
value = [value] * self.k_exog_params
self._params_split['exog_params'] = validate_basic(
value, self.k_exog_params, title='exogenous coefficients')
self._params = None
def sigma2(self, params):
length = int(not self.spec.concentrate_scale)
self._params_split['sigma2'] = validate_basic(params,
length,
title='sigma2').item()
self._params = None
def seasonal_ma_params(self, value):
if np.isscalar(value):
value = [value] * self.k_seasonal_ma_params
self._params_split['seasonal_ma_params'] = validate_basic(
value, self.k_seasonal_ma_params, title='seasonal MA coefficients')
self._params = None