API, list of functions

kronlininv.blockpostcov

Compute a block of the posterior covariance defined by the min/max indices astart/aend (rows), bstart/bend (columns).

Parameters

• factkli – (named tuple): a set of arrays computed with the function calcfactors()

• astart,aend (integers) – the first and last row of the posterior covariance to be computed

• bstart,bend (integers) – the first and last column of the posterior covariance to be computed

Returns

a block of the posterior covariance

Return type

ndarray

kronlininv.calcfactors(G1, G2, G3, Cm1, Cm2, Cm3, Cd1, Cd2, Cd3)

Compute the factors required to solve the inverse problem. These factors are the ones to be stored to compute the posterior covariance or mean.

Parameters

• G1,G2,G3 (ndarrays) – the three forward model matrices

• Cm1,Cm2,Cm3 (ndarrays) – the three covariance matrices related to the model parameters

• Cd1,Cd2,Cd3 (ndarrays) – the three covariance matrices related to the observed data

Returns

a set of arrays required to subsequently calculate the posterior mean or covariance

Return type

named tuple

kronlininv.posteriormean

Compute the posterior mean model.

Parameters

• factkli (named tuple) – a set of arrays computed with the function calcfactors()

• G1,G2,G3 (ndarrays) – the three forward model matrices

• mprior (ndarray, 1D) – the prior model

• dobs (ndarray, 1D) – the observed data

Returns

the posterior mean model

Return type

ndarray, 1D