clophfit.fitting.error_models#

Error models for fitting.

Classes#

ErrorModel

Protocol for error models.

ConstantErrorModel

Homoscedastic error.

ProportionalErrorModel

Simple Heteroscedastic: var = sigma_read^2 + (rel_error * signal)^2.

ComprehensiveErrorModel

Physical error: var = sigma_read^2 + gain * signal + (rel_error * signal)^2.

Module Contents#

class clophfit.fitting.error_models.ErrorModel#

Bases: Protocol

Protocol for error models.

compute_variance(signal, label='')#

Compute variance given a signal estimate.

Parameters:

  • signal (clophfit.clophfit_types.ArrayF)

  • label (int | str)

Return type:

clophfit.clophfit_types.ArrayF

class clophfit.fitting.error_models.ConstantErrorModel(sigma_read)#

Homoscedastic error.

Parameters:

sigma_read (float | ArrayF | Mapping[int | str, float | ArrayF]) – Constant error/variance floor.

compute_variance(signal, label='')#

Compute variance.

Parameters:

  • signal (clophfit.clophfit_types.ArrayF)

  • label (int | str)

Return type:

clophfit.clophfit_types.ArrayF

class clophfit.fitting.error_models.ProportionalErrorModel(sigma_read, rel_error)#

Simple Heteroscedastic: var = sigma_read^2 + (rel_error * signal)^2.

Read noise + proportional noise (ignoring Poisson scaling).

Parameters:

  • sigma_read (float | ArrayF | Mapping[int | str, float | ArrayF]) – Read noise floor.

  • rel_error (float | Mapping[int | str, float]) – Proportional error coefficient.

compute_variance(signal, label='')#

Compute variance.

Parameters:

  • signal (clophfit.clophfit_types.ArrayF)

  • label (int | str)

Return type:

clophfit.clophfit_types.ArrayF

class clophfit.fitting.error_models.ComprehensiveErrorModel(sigma_read, gain, rel_error)#

Physical error: var = sigma_read^2 + gain * signal + (rel_error * signal)^2.

Shot Noise + Proportional: Read noise + Poisson shot noise + Scintillation/Proportional noise.

Parameters:

  • sigma_read (float | ArrayF | Mapping[int | str, float | ArrayF]) – Read noise floor.

  • gain (float | Mapping[int | str, float]) – Instrument gain linking signal to variance (Poisson term).

  • rel_error (float | Mapping[int | str, float]) – Proportional error coefficient.

compute_variance(signal, label='')#

Compute variance.

Parameters:

  • signal (clophfit.clophfit_types.ArrayF)

  • label (int | str)

Return type:

clophfit.clophfit_types.ArrayF