GALAHAD IR package#
purpose#
Given a sparse symmetric \(n \times n\) matrix \(A = a_{ij}\) and the
factorization of \(A\) found by the GALAHAD package SLS, the ir
package
solves the system of linear equations \(A x = b\) using
iterative refinement.
Currently only the options and inform dictionaries are exposed; these are provided and used by other GALAHAD packages with Python interfaces. Please contact us if you would like full functionality!
See Section 4 of $GALAHAD/doc/ir.pdf for additional details.
method#
Iterative refinement proceeds as follows. First obtain the floating-point solution to \(A x = b\) using the factors of \(A\). Then iterate until either the desired residual accuracy (or the iteration limit is reached) as follows: evaluate the residual \(r = b - A x\), find the floating-point solution \(\delta x\) to \(A \delta x = r\), and replace \(x\) by \(x + \delta x\).
parametric real type T#
Below, the symbol T refers to a parametric real type that may be Float32 (single precision), Float64 (double precision) or, if supported, Float128 (quadruple precision).
callable functions#
function ir_initialize(T, data, control, status)
Set default control values and initialize private data
Parameters:
data |
holds private internal data |
control |
is a structure containing control information (see ir_control_type) |
status |
is a scalar variable of type Int32 that gives the exit status from the package. Possible values are (currently):
|
function ir_information(T, data, inform, status)
Provides output information
Parameters:
data |
holds private internal data |
inform |
is a structure containing output information (see ir_inform_type) |
status |
is a scalar variable of type Int32 that gives the exit status from the package. Possible values are (currently):
|
function ir_terminate(T, data, control, inform)
Deallocate all internal private storage
Parameters:
data |
holds private internal data |
control |
is a structure containing control information (see ir_control_type) |
inform |
is a structure containing output information (see ir_inform_type) |
available structures#
ir_control_type structure#
struct ir_control_type{T} f_indexing::Bool error::Int32 out::Int32 print_level::Int32 itref_max::Int32 acceptable_residual_relative::T acceptable_residual_absolute::T required_residual_relative::T record_residuals::Bool space_critical::Bool deallocate_error_fatal::Bool prefix::NTuple{31,Cchar}
detailed documentation#
control derived type as a Julia structure
components#
Bool f_indexing
use C or Fortran sparse matrix indexing
Int32 error
unit for error messages
Int32 out
unit for monitor output
Int32 print_level
controls level of diagnostic output
Int32 itref_max
maximum number of iterative refinements allowed
T acceptable_residual_relative
refinement will cease as soon as the residual \(\|Ax-b\|\) falls below max( acceptable_residual_relative \* \(\|b\|\), acceptable_residual_absolute )
T acceptable_residual_absolute
see acceptable_residual_relative
T required_residual_relative
refinement will be judged to have failed if the residual \(\|Ax-b\| \geq\) required_residual_relative \* \(\|b\|\). No checking if required_residual_relative < 0
Bool record_residuals
record the initial and final residual
Bool space_critical
if space is critical, ensure allocated arrays are no bigger than needed
Bool deallocate_error_fatal
exit if any deallocation fails
NTuple{31,Cchar} prefix
all output lines will be prefixed by prefix(2:LEN(TRIM(.prefix))-1) where prefix contains the required string enclosed in quotes, e.g. “string” or ‘string’
ir_inform_type structure#
struct ir_inform_type{T} status::Int32 alloc_status::Int32 bad_alloc::NTuple{81,Cchar} norm_initial_residual::T norm_final_residual::T
detailed documentation#
inform derived type as a Julia structure
components#
Int32 status
the return status. Possible values are:
0
the solution has been found.
-1
An allocation error occurred. A message indicating the offending array is written on unit control.error, and the returned allocation status and a string containing the name of the offending array are held in inform.alloc_status and inform.bad_alloc respectively.
-2
A deallocation error occurred. A message indicating the offending array is written on unit control.error and the returned allocation status and a string containing the name of the offending array are held in inform.alloc_status and inform.bad_alloc respectively.
-11
Iterative refinement has not reduced the relative residual by more than control.required_relative_residual.
Int32 alloc_status
the status of the last attempted allocation/deallocation.
NTuple{81,Cchar} bad_alloc
the name of the array for which an allocation/deallocation error occurred.
T norm_initial_residual
the infinity norm of the initial residual
T norm_final_residual
the infinity norm of the final residual