jacobianFromYbatch

Workspace.jacobianFromYbatch(self: pyarts.arts._Workspace, jacobian: Optional[Union[pyarts.arts.WorkspaceVariable, pyarts.arts.Matrix]] = self.jacobian, ybatch: Optional[Union[pyarts.arts.WorkspaceVariable, pyarts.arts.ArrayOfVector]] = self.ybatch, y: Optional[Union[pyarts.arts.WorkspaceVariable, pyarts.arts.Vector]] = self.y, pert_size: Union[pyarts.arts.WorkspaceVariable, pyarts.arts.Numeric], verbosity: Optional[Union[pyarts.arts.WorkspaceVariable, pyarts.arts.Verbosity]] = self.verbosity) → None

Sets jacobian based on perturbation calcuations.

This function assumes that ybatch contains spectra calculated with some variable perturbed, in comparison to the calculation behind y. The function takes the differences between ybatch and y to form a numerical derived estimate of jacobian.

Column i of jacobian equals: (ybatch[i]-y)/pert_size.

Author(s): Patrick Eriksson

Parameters:

• jacobian (Matrix, optional) – The Jacobian matrix. See jacobian, defaults to self.jacobian [OUT]

• ybatch (ArrayOfVector, optional) – Batch of spectra. See ybatch, defaults to self.ybatch [IN]

• y (Vector, optional) – The measurement vector. See y, defaults to self.y [IN]

• pert_size (Numeric) – Size of perturbation behind spectra in ybatch. [IN]

• verbosity (Verbosity) – ARTS verbosity. See verbosity, defaults to self.verbosity [IN]