fouriax.SensorNoiseModel#

class SensorNoiseModel#

Bases: ABC

Base interface for stochastic sensor noise applied to expected measurements.

__init__()#

Methods

__init__()

covariance(expected)

Return the analytic measurement covariance matrix.

expected_variance(expected)

Return per-element variance for the noise model at the expected signal.

precision(expected, *[, regularize])

Return the analytic inverse covariance (precision) matrix.

sample(expected, *, key)

Draw a noisy sample from the measurement distribution.

abstractmethod sample(expected, *, key)#

Draw a noisy sample from the measurement distribution.

Parameters:
  • expected (Array)

  • key (Array)

Return type:

Array

abstractmethod expected_variance(expected)#

Return per-element variance for the noise model at the expected signal.

Parameters:

expected (Array)

Return type:

Array

covariance(expected)#

Return the analytic measurement covariance matrix.

The default implementation assumes conditionally independent noise and constructs a diagonal covariance from expected_variance(…).

Parameters:

expected (Array)

Return type:

Array

precision(expected, *, regularize=1e-12)#

Return the analytic inverse covariance (precision) matrix.

The default implementation assumes conditionally independent noise and constructs a diagonal precision from expected_variance(…).

Parameters:
  • expected (Array)

  • regularize (float)

Return type:

Array