def test_interval(self):
"""
Tests :func:`colour.utilities.array.interval` definition.
"""
np.testing.assert_almost_equal(interval(range(0, 10, 2)),
np.array([2]))
np.testing.assert_almost_equal(interval([1, 2, 3, 4, 6, 6.5]),
np.array([0.5, 1, 2]))
def test_interval(self):
"""
Tests :func:`colour.utilities.array.interval` definition.
"""
np.testing.assert_almost_equal(
interval(range(0, 10, 2)),
np.array([2]))
np.testing.assert_almost_equal(
interval([1, 2, 3, 4, 6, 6.5]),
np.array([0.5, 1, 2]))
def x(self, value):
"""
Setter for **self.__x** private attribute.
Parameters
----------
value : array_like
Attribute value.
"""
if value is not None:
value = np.atleast_1d(value).astype(np.float_)
assert value.ndim == 1, (
'"x" independent variable must have exactly one dimension!')
value_interval = interval(value)[0]
xp1 = value[0] - value_interval * 2
xp2 = value[0] - value_interval
xp3 = value[-1] + value_interval
xp4 = value[-1] + value_interval * 2
self._xp = np.concatenate(((xp1, xp2), value, (xp3, xp4)))
self.__x = value
def x(self, value):
"""
Setter for **self.__x** private attribute.
Parameters
----------
value : array_like
Attribute value.
"""
if value is not None:
value = np.atleast_1d(value).astype(np.float_)
assert value.ndim == 1, (
'"x" independent variable must have exactly one dimension!')
value_interval = interval(value)[0]
xp1 = value[0] - value_interval * 2
xp2 = value[0] - value_interval
xp3 = value[-1] + value_interval
xp4 = value[-1] + value_interval * 2
self._xp = np.concatenate(((xp1, xp2), value, (xp3, xp4)))
self.__x = value
def _evaluate(self, x):
"""
Performs the interpolator evaluation at given points.
Parameters
----------
x : ndarray
Points to evaluate the interpolant at.
Returns
-------
ndarray
Interpolated points values.
"""
self._validate_dimensions()
self._validate_interpolation_range(x)
x_interval = interval(self._x)[0]
x_f = np.floor(x / x_interval)
windows = (x_f[:, np.newaxis] + np.arange(-self._window + 1,
self._window + 1))
clip_l = min(self._x_p) / x_interval
clip_h = max(self._x_p) / x_interval
windows = np.clip(windows, clip_l, clip_h) - clip_l
windows = np.around(windows).astype(np.int_)
return np.sum(
self._y_p[windows] *
self._kernel(x[:, np.newaxis] / x_interval - windows -
min(self._x_p) / x_interval, **self._kernel_args),
axis=-1)
def x(self, value):
"""
Setter for the **self.x** property.
"""
if value is not None:
value = np.atleast_1d(value).astype(self._dtype)
assert value.ndim == 1, (
'"x" independent variable must have exactly one dimension!')
value_interval = interval(value)
if value_interval.size != 1:
warning(('"x" independent variable is not uniform, '
'unpredictable results may occur!'))
self._x = value
if self._window is not None:
self._x_p = np.pad(
self._x, (self._window, self._window),
'linear_ramp',
end_values=(
np.min(self._x) - self._window * value_interval[0],
np.max(self._x) + self._window * value_interval[0]))
def x(self, value):
"""
Setter for the **self.x** property.
"""
if value is not None:
value = np.atleast_1d(value).astype(self._dtype)
assert value.ndim == 1, (
'"x" independent variable must have exactly one dimension!')
value_interval = interval(value)[0]
xp1 = value[0] - value_interval * 2
xp2 = value[0] - value_interval
xp3 = value[-1] + value_interval
xp4 = value[-1] + value_interval * 2
self._xp = np.concatenate(((xp1, xp2), value, (xp3, xp4)))
self._x = value