STPSV/DTPSV/CTPSV/ZTPSV

 

Solve triangular system

 

trans

Transposition option for A:

 

 

’N’ or ’n’

Compute x ← A−1x

 

 

’T’ or ’t’

Compute x ← A−Tx

 

 

’C’ or ’c’

Compute x ← A-*x

 

 

where A−T is the inverse of the transpose of A, and A–*

 

 

is the inverse of the conjugate transpose. In the real

 

 

subprograms, ’C’ and ’c’ have the same meaning as ’T’

 

 

and ’t’.

 

 

diag

Specifies whether the matrix is unit triangular, that is,

 

 

aii = 1, or not:

 

 

 

’N’ or ’n’

The diagonal of A is stored in the

 

 

 

array

 

 

’U’ or ’u’

The diagonal of A consists of unstored

 

 

 

ones

 

 

When diag is supplied as ’U’ or ’u’, the diagonal

 

 

elements are not referenced.

 

n

Number of rows and columns in matrix A, n ≥ 0. If

 

 

n = 0, the subprograms do not reference ap or x.

 

ap

Array of length lenap = n⋅(n+1)/2 containing the

 

 

n-by-ntriangular matrix A, stored by columns in the

 

 

packed form described above. Space must be left for the

 

 

diagonal elements of A even when diag is supplied as

 

 

’U’ or ’u’.

 

 

x

Array of length lenx = (n−1)⋅incx+1 containing the

 

 

right-hand-side n-vector x.

 

incx

Increment for the array x, incx ≠ 0:

 

 

incx > 0

x is stored forward in array x; that is,

 

 

 

xi is stored in x((i−1)⋅incx+1)

 

 

incx < 0

x is stored backward in array x; that

 

 

 

is, xi is stored in x((i−n)⋅incx+1)

 

 

Use incx = 1 if the vector x is stored contiguously in

 

 

array x, that is, if xi is stored in x(i). Refer to “BLAS

 

 

Indexing Conventions” in the introduction to

 

 

Chapter 2.

 

Output

x

The solution vector of the triangular system replaces

 

 

the input.

 

316HP MLIB User’s Guide