pastas.noisemodels.ArmaNoiseModel

class pastas.noisemodels.ArmaNoiseModel(model: pastas.typing.Model, name: str = 'noise', norm: bool = True)

ARMA(1,1) Noise model to simulate the noise as defined in Collenteur et al. [2021].

Warning

This model has only been tested on regular time steps and should not be used for irregular time steps yet.

Notes

Calculates the noise according to:

\[\begin{split}\\upsilon_t = r_t - r_{t-1} e^{-\\Delta t/\\alpha} - \\upsilon_{t-1} e^{-\\Delta t/\\beta}\end{split}\]

The units of the alpha and beta parameters are always in days.

property nparam: int

Number of parameters for the noise model.

Returns:

Number of parameters (2 for ArmaNoiseModel: alpha and beta).

Return type:

int

Methods

set_init_parameters(→ None)	Set initial parameters for the noise model.
simulate(→ pandas.Series)	Simulate noise from the residual series.
calculate_noise(→ pastas.typing.ArrayLike)	Calculate the noise values for the ARMA(1,1) noise model.
to_dict(→ dict)	Return a dict to store the noise model.