CCHKHB2STG(3)
π CCHKHB2STG(3) π
Library Functions Manual
Okay bestie, buckle up β this is the LAPACK test-driver page for CCHKHB2STG, and we are gonna make a Fortran subroutine feel like a group chat. Every flag, dimension, and slightly cursed blank spot is staying exactly where it belongs. No cap. π
π NAME π
/home/abuild/rpmbuild/BUILD/lapack-static-3.12.0-build/lapack-3.12.0/TESTING/EIG/cchkhb2stg.f
That's the whole NAME entry β it's giving "file path as a name" energy, and honestly? Iconic. π
π¦ SYNOPSIS π¦
π§ Functions/Subroutines π§
subroutine CCHKHB2STG (nsizes, nn, nwdths, kk, ntypes, dotype, iseed,
thresh, nounit, a, lda, sd, se, d1, d2, d3, u, ldu, work, lwork,
rwork, result, info)
CCHKHB2STG
π«Ά Function/Subroutine Documentation π«Ά
Here's the full typed-out signature, bestie β yes, it's a whole paragraph and yes, we're respecting every character:
subroutine CCHKHB2STG (integer nsizes, integer, dimension( * ) nn, integer
nwdths, integer, dimension( * ) kk, integer ntypes, logical, dimension(
* ) dotype, integer, dimension( 4 ) iseed, real thresh, integer nounit,
complex, dimension( lda, * ) a, integer lda, real, dimension( * ) sd,
real, dimension( * ) se, real, dimension( * ) d1, real, dimension( * )
d2, real, dimension( * ) d3, complex, dimension( ldu, * ) u, integer
ldu, complex, dimension( * ) work, integer lwork, real, dimension( * )
rwork, real, dimension( * ) result, integer info)
CCHKHB2STG
CCHKHB2STG
π Purpose π
CCHKHB2STG tests the reduction of a Hermitian band matrix to tridiagonal from, used with the Hermitian eigenvalue problem. (Yes, "from" is literally what the page says β I'm not editing the lore. π)
CHBTRD factors a Hermitian band matrix A as U S U*, where * means conjugate transpose, S is symmetric tridiagonal, and U is unitary. CHBTRD can use either just the lower or just the upper triangle of A; CCHKHB2STG checks both cases. Both triangles get their moment. π
CHETRD_HB2ST factors a Hermitian band matrix A as U S U*, where * means conjugate transpose, S is symmetric tridiagonal, and U is unitary. CHETRD_HB2ST can use either just the lower or just the upper triangle of A; CCHKHB2STG checks both cases.
DSTEQR factors S as Z D1 Z'. D1 is the matrix of eigenvalues computed when Z is not computed and from the S resulting of DSBTRD (used as reference for DSYTRD_SB2ST). D2 is the matrix of eigenvalues computed when Z is not computed and from the S resulting of DSYTRD_SB2ST. D3 is the matrix of eigenvalues computed when Z is not computed and from the S resulting of DSYTRD_SB2ST.
When CCHKHB2STG is called, a number of matrix (), a number of bandwidths (), and a number of matrix are specified. For each size (), each bandwidth () less than or equal to , and each type of matrix, one matrix will be generated and used to test the hermitian banded reduction routine. For each matrix, a number of tests will be performed:
(1) | A - V S V* | / ( |A| n ulp ) computed by CHBTRD with
UPLO='U'
(2) | I - UU* | / ( n ulp )
(3) | A - V S V* | / ( |A| n ulp ) computed by CHBTRD with
UPLO='L'
(4) | I - UU* | / ( n ulp )
(5) | D1 - D2 | / ( |D1| ulp ) where D1 is computed by
DSBTRD with UPLO='U' and
D2 is computed by
CHETRD_HB2ST with UPLO='U'
(6) | D1 - D3 | / ( |D1| ulp ) where D1 is computed by
DSBTRD with UPLO='U' and
D3 is computed by
CHETRD_HB2ST with UPLO='L'
The are specified by an array NN(1:NSIZES); the value of each element NN(j) specifies one size. The are specified by a logical array DOTYPE( 1:NTYPES ); if DOTYPE(j) is .TRUE., then matrix type will be generated. Currently, the list of possible types is:
(1)The zero matrix. β minimalism, we love her.(2)The identity matrix. β main character energy.(3)A diagonal matrix with evenly spaced entries1, ..., ULPand random signs. (ULP = (first number larger than 1) - 1) β the math is right there.(4)A diagonal matrix with geometrically spaced entries1, ..., ULPand random signs. β geometry era.(5)A diagonal matrix with entries1, ULP, ..., ULPand random signs.(6)Same as(4), but multiplied bySQRT( overflow threshold ). β spicy scaling.(7)Same as(4), but multiplied bySQRT( underflow threshold ). β tiny scaling.(8)A matrix of the formU* D U, whereUis unitary andDhas evenly spaced entries1, ..., ULPwith random signs on the diagonal.(9)A matrix of the formU* D U, whereUis unitary andDhas geometrically spaced entries1, ..., ULPwith random signs on the diagonal.(10)A matrix of the formU* D U, whereUis unitary andDhas entries1, ULP,..., ULPwith random signs on the diagonal.(11)Same as(8), but multiplied bySQRT( overflow threshold ).(12)Same as(8), but multiplied bySQRT( underflow threshold ).(13)Hermitian matrix with random entries chosen from(-1,1).(14)Same as(13), but multiplied bySQRT( overflow threshold ).(15)Same as(13), but multiplied bySQRT( underflow threshold ).
π Parameters π
Okay bestie, parameter roll call β every single one, no skips:
NSIZESβNSIZES is INTEGER. The number of sizes of matrices to use. If it is zero,CCHKHB2STGdoes nothing. It must be at least zero.NNβNN is INTEGER array, dimension (NSIZES). An array containing the sizes to be used for the matrices. Zero values will be skipped. The values must be at least zero.NWDTHSβNWDTHS is INTEGER. The number of bandwidths to use. If it is zero,CCHKHB2STGdoes nothing. It must be at least zero.KKβKK is INTEGER array, dimension (NWDTHS). An array containing the bandwidths to be used for the band matrices. The values must be at least zero.NTYPESβNTYPES is INTEGER. The number of elements inDOTYPE. If it is zero,CCHKHB2STGdoes nothing. It must be at least zero. If it isMAXTYP+1andNSIZESis 1, then an additional type,MAXTYP+1is defined, which is to use whatever matrix is inA. This is only useful ifDOTYPE(1:MAXTYP)is.FALSE.andDOTYPE(MAXTYP+1)is.TRUE..DOTYPEβDOTYPE is LOGICAL array, dimension (NTYPES). IfDOTYPE(j)is.TRUE., then for each size inNNa matrix of that size and of typejwill be generated. IfNTYPESis smaller than the maximum number of types defined (PARAMETER MAXTYP), then typesNTYPES+1throughMAXTYPwill not be generated. IfNTYPESis larger thanMAXTYP,DOTYPE(MAXTYP+1)throughDOTYPE(NTYPES)will be ignored.ISEEDβISEED is INTEGER array, dimension (4). On entryISEEDspecifies the seed of the random number generator. The array elements should be between 0 and 4095; if not they will be reduced mod 4096. Also,ISEED(4)must be odd. The random number generator uses a linear congruential sequence limited to small integers, and so should produce machine independent random numbers. The values ofISEEDare changed on exit, and can be used in the next call toCCHKHB2STGto continue the same random number sequence.THRESHβTHRESH is REAL. A test will count as if the , computed as described above, exceedsTHRESH. Note that the error is scaled to be O(1), soTHRESHshould be a reasonably small multiple of 1, e.g., 10 or 100. In particular, it should not depend on the precision (single vs. double) or the size of the matrix. It must be at least zero.NOUNITβNOUNIT is INTEGER. The FORTRAN unit number for printing out error messages (e.g., if a routine returnsIINFOnot equal to 0.)AβA is COMPLEX array, dimension (LDA, max(NN)). Used to hold the matrix whose eigenvalues are to be computed.LDAβLDA is INTEGER. The leading dimension ofA. It must be at least 2 (not 1!) and at leastmax( KK )+1.SDβSD is REAL array, dimension (max(NN)). Used to hold the diagonal of the tridiagonal matrix computed byCHBTRD.SEβSE is REAL array, dimension (max(NN)). Used to hold the off-diagonal of the tridiagonal matrix computed byCHBTRD.D1βD1 is REAL array, dimension (max(NN)). Used store eigenvalues resulting from the tridiagonal form using theDSBTRD.D2βD2 is REAL array, dimension (max(NN)).D3βD3 is REAL array, dimension (max(NN)).UβU is COMPLEX array, dimension (LDU, max(NN)). Used to hold the unitary matrix computed byCHBTRD.LDUβLDU is INTEGER. The leading dimension ofU. It must be at least 1 and at leastmax( NN ).WORKβWORK is COMPLEX array, dimension (LWORK).LWORKβLWORK is INTEGER. The number of entries inWORK. This must be at leastmax( LDA+1, max(NN)+1 )*max(NN).RWORKβRWORK is REAL array.RESULTβRESULT is REAL array, dimension (4). The values computed by the tests described above. The values are currently limited to1/ulp, to avoid overflow.INFO
INFO is INTEGER β if it's 0, then everything ran OK. That's the green flag, bestie. π’β¨
π§Έ Some Local Variables and Parameters
ZERO,ONEβ real0and1, the iconic little constants.MAXTYPβ the number of types defined.NTESTβ the number of tests performed, or which can be performed so far, for the current matrix.NTESTTβ the total number of tests performed so far.NMAXβ largest value inNN.NMATSβ the number of matrices generated so far.NERRSβ the number of tests which have exceededTHRESHso far.COND,IMODEβ values to be passed to the matrix generators.ANORMβ norm ofA; passed to matrix generators.OVFL,UNFLβ overflow and underflow thresholds.ULP,ULPINVβ finest relative precision and its inverse.RTOVFL,RTUNFLβ square roots of the previous 2 values.
The following four arrays decode JTYPE:
KTYPE(j)β the general type (1-10) for type .KMODE(j)β theMODEvalue to be passed to the matrix generator for type .KMAGN(j)β the order of magnitude (O(1),O(overflow^(1/2) ),O(underflow^(1/2) )).
π Author
- Univ. of Tennessee
- Univ. of California Berkeley
- Univ. of Colorado Denver
- NAG Ltd.
Definition at line 337 of file cchkhb2stg.f.
π Author
Generated automatically by Doxygen for LAPACK from the source code.
That's the whole vibe from LAPACK Version 3.12.0, page /home/abui...chkhb2stg.f(3) β the footer didn't hand me a date, so I'm not making one up. Slay responsibly. π
β¨
π¦ where this page comes from
- opensuse-leap 16.0: lapack-man 3.12.0-160000.2.2 Β· BSD-3-Clause
- opensuse-leap 16.0: lapack-man 3.12.0-160000.2.2 Β· BSD-3-Clause
- opensuse-leap 16.0: lapack-man 3.12.0-160000.2.2 Β· BSD-3-Clause
- opensuse-leap 16.0: lapack-man 3.12.0-160000.2.2 Β· BSD-3-Clause
- π girlified by deepseek-ai/DeepSeek-V4.1-Flash-1017 on : a language model's rewrite of the original, under its terms