incline.Heteroskedastic

class incline.Heteroskedastic(standard_deviation=None, window_length=25)[source]

Independent noise whose scale changes across the series.

A constant noise model cannot represent changing local scale: it tends to overstate uncertainty in quieter regions and understate it in noisier ones.

Variables:
  • standard_deviation (numpy._typing._array_like.NDArray[numpy.float64] | None) – Per-point standard deviations. Estimated locally when None.

  • window_length (int) – Points averaged by the local estimator.

Parameters:
  • standard_deviation (NDArray[float64] | None)

  • window_length (int)

__init__(standard_deviation=None, window_length=25)
Parameters:
  • standard_deviation (NDArray[float64] | None)

  • window_length (int)

Return type:

None

Methods

__init__([standard_deviation, window_length])

estimate(y, axis)

Estimate the local noise level, or use the supplied one.

Attributes

standard_deviation: NDArray[float64] | None = None
window_length: int = 25
estimate(y, axis)[source]

Estimate the local noise level, or use the supplied one.

Parameters:
  • y (npt.NDArray[np.float64])

  • axis (TimeAxis)

Return type:

NoiseFit