ARPACK provides a means to trace the progress of the computation
as it proceeds. Various levels of output may be specified
from no output, level = 0, to voluminous, level = 3.
The following statements may be used within the calling program to
initiate and request this output.
include 'debug.h'
ndigit = -3
logfil = 6
msgets = 0
msaitr = 0
msapps = 0
msaupd = 1
msaup2 = 0
mseigt = 0
mseupd = 0
The parameter "logfil" specifies the logical unit number of the output
file. The parameter "ndigit" specifies the number of decimal digits
and the width of the output lines. A positive value of "ndigit"
specifies that 132 columns are used during output and a negative
value specifies eighty columns are to be used. The values of the remaining
parameters indicate the output levels from the indicated routines.
For the above example, "msaitr" indicates the level of output requested
for the subroutine ssaitr or dsaitr. The above configuration will
give a breakdown of the number of matrix vector products required,
the total number of iterations, the number of re-orthogonalization
steps and an estimate of the time spent in each routine and phase of the
computation. The following output is produced:
---------------------------------------------------------------------
==========================================
= Symmetric implicit Arnoldi update code =
= Version Number: 2.1 =
= Version Date: 11/15/95 =
==========================================
= Summary of timing statistics =
==========================================
Total number update iterations = 8
Total number of OP*x operations = 125
Total number of B*x operations = 0
Total number of reorthogonalization steps = 125
Total number of iterative refinement steps = 0
Total number of restart steps = 0
Total time in user OP*x operation = 0.020002
Total time in user B*x operation = 0.000000
Total time in Arnoldi update routine = 0.210021
Total time in ssaup2 routine = 0.190019
Total time in basic Arnoldi iteration loop = 0.110011
Total time in reorthogonalization phase = 0.070007
Total time in (re)start vector generation = 0.000000
Total time in trid eigenvalue subproblem = 0.040004
Total time in getting the shifts = 0.000000
Total time in applying the shifts = 0.040004
Total time in convergence testing = 0.000000
---------------------------------------------------------------------
The user is encouraged to experiment with the other settings
once some familiarity has been gained with the routines.
The include statement sets up the storage declarations that are
solely associated with this trace debugging feature. "debug.h"
has the following structure:
---------------------------------------------------------------------
c
c\SCCS Information: @(#)
c FILE: debug.h SID: 2.3 DATE OF SID: 11/16/95 RELEASE: 2
c
c %---------------------------------%
c | See debug.doc for documentation |
c %---------------------------------%
integer logfil, ndigit, mgetv0,
& msaupd, msaup2, msaitr, mseigt, msapps, msgets, mseupd,
& mnaupd, mnaup2, mnaitr, mneigh, mnapps, mngets, mneupd,
& mcaupd, mcaup2, mcaitr, mceigh, mcapps, mcgets, mceupd
common /debug/
& logfil, ndigit, mgetv0,
& msaupd, msaup2, msaitr, mseigt, msapps, msgets, mseupd,
& mnaupd, mnaup2, mnaitr, mneigh, mnapps, mngets, mneupd,
& mcaupd, mcaup2, mcaitr, mceigh, mcapps, mcgets, mceupd
---------------------------------------------------------------------
The parameters "msaupd, msaup2, msaitr, mseigt, msapps, msgets, mseupd"
are for the symmetric codes, while
"mnaupd, mnaup2, mnaitr, mneigh, mnapps, mngets, mneupd" are for the
nonsymmetric codes and, finally,
"mcaupd, mcaup2, mcaitr, mceigh, mcapps, mcgets, mceupd" are for the complex
arithmetic codes. A comprehensive break down of each parameter is given
below.
==========================================================
=== Common to symmetric, nonsymmetric and complex code ===
==========================================================
logfil: unit number where the logfile (debug) is written
ndigit: number of digits used in the debug output
ndigit < 0: printing is done with 72 columns.
ndigit > 0: printing is done with 132 columns.
mgetv0 > 0: print residual vector generated.
======================================
=== Specific to the symmetric code ===
======================================
msaupd > 0: *Print the number of iterations taken,
number of "converged" eigenvalues,
final Ritz values and corresponding Ritz estimates.
*Print various timing statistics.
msaup2 > 0: *Print major iteration number,
number of "converged" Ritz values on exit,
B-norm of the residual vector of length NCV factorization,
B-norm of the residual vector of length NEV factorization,
residual norm before exit,
Ritz values and corresponding Ritz estimates before exit.
msaup2 > 1: print number of unreduced submatrices,
Ritz values and corresponding Ritz estimates of the current
T matrix, actual values for NEV and NP,
wanted Ritz values and corresponding Ritz estimates,
shifts selected.
msaup2 > 2: print "unwanted" Ritz values and corresponding Ritz
estimates, order NCV matrix T (diagonal and off-diagonal),
unwanted Ritz values and error bounds.
msaitr > 0: print iteration number, residual norm, restart info
print if an off diagonal element of T became negative.
msaitr > 1: print the final matrix T.
msaitr > 2: print Arnoldi vector no. generate at iteration j,
b-norm of residual vector at each iteration,
print rnorm and rnorm1 for iterative refinement,
print wnorm and rnorm used for Re-orthogonalization,
V^T * B * (resid/B-norm(resid)),
print the results of whether the current residual vector is
orthogonal to the current Lanczos basis.
msaitr > 3: print the matrix T at each iteration.
print the residual vector and arnoldi vectors.
mseigt > 0: print the current matrix T.
msgets > 0: print NEV and NP,
eigenvalues of and corresponding Ritz estimates of the
current T matrix.
msapps > 0: print information about deflation at row/column no.
msapps > 1: print initial matrix T
print sigmak, betak and matrix T after all shifts
msapps > 2: print the matrix T after the application of each shift.
msapps > 3: updated residual for next iteration.
mseupd > 1: print eigenvalues of the final T matrix,
the last row of the eigenvector matrix for T,
if reordered, reordered last row of the eigenvector matrix,
reordered NCV Ritz values of the final T matrix,
if type = 'REGULAR', untransformed "converged" Ritz values
and corresponding Ritz estimates,
NCV Ritz values of the final T matrix,
last row of the eigenvector matrix for T,
if reordered, reordered last row of the eigenvector matrix,
reordered NCV Ritz values of the final T.
mseupd > 2: print the matrix T.
=========================================
=== Specific to the nonsymmetric code ===
=========================================
mnaupd > 0: *Print the number of iterations taken,
number of "converged" eigenvalues,
real and imaginary parts of the converged Ritz values
and their corresponding Ritz estimates,
*Print various timing statistics.
mnaup2 > 0: *Print major iteration number.
*Print the number of "converged" Ritz values on exit,
and the real and imaginary parts of the "converged" Ritz
values and corresponding Ritz estimates.
mnaup2 > 1: *Print the length of the Arnoldi Factorization,
and the B-norm of its residual vector.
*Print NEV and NP, real and imaginary parts of the "wanted"
Ritz values and associated Ritz estimates at each
iteration.
*Print the B-norm of the residual of the compressed
factorization and the compressed upper Hessenberg matrix H.
mnaup2 > 2: *Print the real and imaginary parts of all the Ritz values
and associated Ritz estimates, NEV, NP, NUMCNV, NCONV.
*Print the real and imaginary parts of the shifts. If the
exact shift strategy is used, print the associated Ritz
estimates of the shifts.
*Print the real and imaginary parts of the Ritz values
and the corresponding Ritz estimates obtained from _neigh.
mnaitr > 0: *Print if a restart is needed.
mnaitr > 1: *Print the number of Arnoldi vector being generated and
the B-norm of the current residual.
mnaitr > 2: *Print j-th column of the Hessenberg matrix H.
*Print reorthogonalization and iterative refinement information,
*Print the final upper Hessenberg matrix of order K+NEV.
mnaitr > 3: *Print V^T*B*resid/(B-norm(resid)).
mnaitr > 4: *Print current upper Hessenberg matrix.
mnaitr > 5: *Print updated arnoldi vectors and the residual vector.
mneigh > 1: *Print the last row of the Schur matrix for H, and
the last row of the eigenvector matrix for H.
mneigh > 2: *Print the entering upper Hessenberg matrix.
*Print the real and imaginary part of eigenvalues
of the current Hessenberg matrix, and associated
Ritz estimates.
mngets > 0: *Print the real and imaginary parts of the Ritz values
of the Hessenberg matrix and their the corresponding
error bounds, KEV, NP.
mnapps > 0: *Print information about where deflation occured.
mnapps > 1: *Print sigmak, betak, order of the final Hessenberg matrix,
and the final compressed upper Hessenberg matrix.
mnapps > 2: *Print implicit application of shift number, real and imaginary
part of the shift.
*Print the indices of the submatrix that the shift is applied.
mnapps > 3: *Print the matrix H before and after the application of
each shift, updated residual for next iteration.
mnapps > 4: *Print the accumulated orthogonal Hessenberg matrix Q,
updated matrix of Arnoldi vectors.
mneupd > 0: *Print the number of converged Ritz values, B-norm of the
residual, all NCV Ritz values and error bounds.
mneupd > 1: *Print the final upper Hessenberg matrix computed by _naupd.
*If Ritz vectors are requested, print real and imaginary parts
of the eigenvalues and the last row of the Schur vectors as
computed by _neupd.
mneupd > 2: *If Ritz vectors are requested, print the threshold eigenvalue
used for re-ordering.
*If Ritz vectors are requested, print the number of eigenvalues
to reorder and the number of converged Ritz values.
*If Ritz vectors are requested, print the upper quasi-matrix
computed by _neupd.
*If Ritz vectors are requested, print the real and imaginary
part of the Ritz values.
*If Ritz vectors are requested, print the last row of the
eigenvector matrix.
*Print the NCV Ritz estimates in the original system.
mneupd > 3: *Print the integer array of pointers.
*If Ritz vectors are requested, print the eigenvector matrix.
*If Ritz vectors are requested, print the reordered upper
quasi-triangular matrix.
mneupd > 4: *If Ritz vectors are requested, print the Q matrix of the QR
factorization of the matrix representing the wanted invariant
subspace.
*If Ritz vectors are requested, print the Schur vectors.
*If Ritz vectors are requested, print the reordered Schur vectors.
====================================
=== Specific to the complex code ===
====================================
mcaupd > 0: *Print the number of iterations taken,
number of "converged" eigenvalues, the converged Ritz values
and their corresponding Ritz estimates,
*Print various timing statistics.
mcaup2 > 0: *Print major iteration number.
*Print the number of "converged" Ritz values on exit, and the
"converged" Ritz values and corresponding Ritz estimates.
mcaup2 > 1: *Print the length of the Arnoldi Factorization,
and the B-norm of its residual vector.
*Print NEV and NP, the "wanted" Ritz values and associated Ritz
estimates at each iteration.
*Print the B-norm of the residual of the compressed
factorization and the compressed upper Hessenberg matrix H.
mcaup2 > 2: *Print the all the Ritz values and associated Ritz estimates,
NEV, NP, NUMCNV, NCONV.
*Print the shifts. If the exact shift strategy is used, print the
associated Ritz estimates of the shifts.
*Print the Ritz values and the corresponding Ritz estimates obtained
from _neigh.
mcaitr > 0: *Print if a restart is needed.
mcaitr > 1: *Print the number of Arnoldi vector being generated and
the B-norm of the current residual.
mcaitr > 2: *Print j-th column of the Hessenberg matrix H.
*Print reorthogonalization and iterative refinement information,
*Print the final upper Hessenberg matrix of order K+NEV.
mcaitr > 3: *Print V^T*B*resid/(B-norm(resid)).
mcaitr > 4: *Print current upper Hessenberg matrix.
mcaitr > 5: *Print updated Arnoldi vectors and the residual vector.
mceigh > 1: *Print the last row of the Schur matrix for H, and
the last row of the eigenvector matrix for H.
mceigh > 2: *Print the entering upper Hessenberg matrix.
*Print the eigenvalues of the current Hessenberg matrix, and
associated Ritz estimates.
mcgets > 0: *Print the real and imaginary parts of the Ritz values
of the Hessenberg matrix and their the corresponding
error bounds, KEV, NP.
mcapps > 0: *Print information about where deflation occured.
mcapps > 1: *Print sigmak, betak, order of the final Hessenberg matrix,
and the final compressed upper Hessenberg matrix.
mcapps > 2: *Print implicit application of shift number, the shift.
*Print the indices of the submatrix that the shift is applied.
mcapps > 3: *Print the matrix H before and after the application of
each shift, updated residual for next iteration.
mcapps > 4: *Print the accumulated unitary Hessenberg matrix Q, and the
updated matrix of Arnoldi vectors.
mceupd > 0: *Print the number of converged Ritz values, B-norm of the
residual, all NCV Ritz values and error bounds.
mceupd > 1: *Print the final upper Hessenberg matrix computed by _naupd.
*If Ritz vectors are requested, print the eigenvalues and the
last row of the Schur vectors as computed by _neupd.
mceupd > 2: *If Ritz vectors are requested, print the threshold eigenvalue
used for re-ordering.
*If Ritz vectors are requested, print the number of eigenvalues
to reorder and the number of converged Ritz values.
*If Ritz vectors are requested, print the upper quasi-matrix
computed by _neupd.
*If Ritz vectors are requested, print the Ritz values.
*If Ritz vectors are requested, print the last row of the
eigenvector matrix.
*Print the NCV Ritz estimates in the original system.
mceupd > 3: *Print the integer array of pointers.
*If Ritz vectors are requested, print the eigenvector matrix.
mceupd > 4: *If Ritz vectors are requested, print the Q matrix of the QR
factorization of the matrix representing the wanted invariant
subspace.
*If Ritz vectors are requested, print the Schur vectors.
