measurement_vector_fitted

Workspace.measurement_vector_fitted: Vector

As measurement_vector, but fitted to the model.

This must often be the same size as measurement_sensor.

See measurement_vector for more details. In that notation, and in the notation of OEM(), \(\vec{y}_f \approx \vec{y} - \epsilon\). Or at least this should be the case depending on how good of a fit of \(\vec{x}\) has been produced and if the measurement can be understood properly.

Tip

It is often useful to present \(\vec{y} - \vec{y}_\epsilon\) and \(\vec{y}_f - \vec{y}_\epsilon\) instead of \(\vec{y}_f\) and \(\vec{y}\) directly. This removes the known measurement error from both the data and the fit, showing the physical signal from the target rather than known sensor noise.

Default value

[]

Modified by workspace method

OEM()

Output from workspace methods

inversion_iterate_agendaExecute() inversion_iterate_agendaExecuteOperator() measurement_inversion_agendaExecute()

measurement_inversion_agendaExecuteOperator() measurement_vector_fittedFromMeasurement()

Output from workspace agendas

inversion_iterate_agenda measurement_inversion_agenda

Related workspace variables

measurement_averaging_kernel measurement_gain_matrix measurement_inversion_agenda measurement_jacobian measurement_jacobian_error

measurement_sensor measurement_vector measurement_vector_error measurement_vector_error_covariance_matrix