The JD/Davidson implementation is focused in the class
Implement a JD_SymSolverBase base class
it also has the functions eigenvalues() and eigenvectors().
The base class also has setters for settings : max_searchspace,num_iterations, convergence_criteria, ...
JDSymSolverBase implements a compute(neigen, target) method, which returns the number of converged eigenvalues and stores the results in eigenvalues and eigenvectors_ member variables.
It also sets the flag info_ which can be accessed by the info() getter, which informs about the success of the calculation. It returns a Specta::CompuationalInfo enum.
The input parameters for the compute method are, number_of_eigenvalues and the Spectra::SortRule
it has a pure virtual function, to compute the correction vector called get_correction_vector
The JD/Davidson implementation is focused in the class
Implement a JD_SymSolverBase base class
compute(neigen, target)
method, which returns the number of converged eigenvalues and stores the results in eigenvalues and eigenvectors_ member variables.