Methods for analyzing the results of electronic structure calculations.
Compute The Pivoted Cholesky Decomposition of a Hermitian Semi-Definite matrix. This is one way to generate localized orbitals.
| Type | Intent | Optional | Attributes | Name | ||
|---|---|---|---|---|---|---|
| type(Matrix_ps), | intent(in) | :: | AMat |
The matrix A, must be hermitian, positive semi-definite. |
||
| type(Matrix_ps), | intent(inout) | :: | LMat |
The matrix computed. |
||
| integer, | intent(in) | :: | rank_in |
The target rank of the matrix. |
||
| type(SolverParameters_t), | intent(in), | optional | :: | solver_parameters_in |
Tarameters for the solver |
When we want to only compute the first n eigenvalues of a matrix, this routine will project out the higher eigenvalues.
| Type | Intent | Optional | Attributes | Name | ||
|---|---|---|---|---|---|---|
| type(Matrix_ps), | intent(inout) | :: | this |
The starting matrix. |
||
| integer, | intent(in) | :: | dim |
The number of eigenvalues ot keep. |
||
| type(Matrix_ps), | intent(inout) | :: | ReducedMat |
a dimxdim matrix with the same first n eigenvalues as the first. |
||
| type(SolverParameters_t), | intent(in), | optional | :: | solver_parameters_in |
The solver parameters. |