Running SIPPI

This section contains information about how to use and control SIPPI, which requires one to

• Define the prior model, $\rho(\mathbf{m})$,in form of the prior data structure

  This will allow sampling from a variety of geostatistical models, using a unified interface, through ´sippi_prior.m´

• Define the forward model, $g(\mathbf{m})$, in form of the forward data structure, and the sippi_forward.m m-file

  This will allow solving a 'forward' problem, such as computhing a geophysical response from some model.

• Define the data and noise model, $L(g(\mathbf{m})$ ,in form of the prior data structure

  This allow describing data and associated uncertainty.

Once these three data structures has been defined, one can solve the associated inverse problem by sampling the posterior distribution, $\rho(\mathbf{m})$.