Compressed Sparse Column (CSC) Format

This standard data format consists of the following data:

integer                   :: m      ! number of rows (unsymmetric only)
integer                   :: n      ! number of columns
integer, size(n+1)        :: ptr    ! column pointers (may have type long)
integer, size(ptr(n+1)-1) :: row    ! row indices
real,    size(ptr(n+1)-1) :: val    ! numerical values

Non-zero matrix entries are ordered by increasing column index and stored in the arrays row(:) and val(:) such that row(k) holds the row number and val(k) holds the value of the k-th entry. The ptr(:) array stores column pointers such that ptr(i) is the position in row(:) and val(:) of the first entry in the i-th column, and ptr(n+1) is one more than the total number of entries. ptr(:) may be either 32-bit (Fortran default integer) or 64-bit (Fortran integer(long)). There must be no duplicate or out of range entries. Entries that are zero, including those on the diagonal, need not be specified.

For symmetric matrices, only the lower triangular entries of \(A\) should be supplied. For unsymmetric matrices, all entries in the matrix should be supplied.

Note that most SPRAL routines offer no checking of user data, and the behaviour of these routines with misformatted data is undefined. You may use routines from the spral_matrix_util package to convert data to and check data stored in this format.

To illustrate the CSC format, the matrix

\[\begin{split}\left( \begin{array}{ccccc} 1.1 & 2.2 & & 3.3 & \\ 2.2 & & 4.4 & & \\ & 4.4 & 5.5 & & 6.6 \\ 3.3 & & & 7.7 & 8.8 \\ & & 6.6 & 8.8 & 9.9 \end{array} \right)\end{split}\]

is described by the following data:

n = 5
ptr(1:6) = (/ 1,             4,   5,        7,        9,    10 /)
row(1:9) = (/ 1,   2,   4,   3,   3,   5,   4,   5,   5 /)
val(1:9) = (/ 1.1, 2.2, 3.3, 4.4, 5.5, 6.6, 7.7, 8.8, 9.9 /)