1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
|
c----------------------------------------------------------------------c
c c
c SPARSKIT Modules FORMATS and UNARY c
c c
c----------------------------------------------------------------------c
c c
c This directory contains both the module FORMATS and UNARY c
c of SPARSKIT. c
c c
c----------------------------------------------------------------------c
c CONTENTS: c
c======================================================================c
c c
c formats.f : mostly format conversion routines. c
c c
c unary.f : elementary manipulation routines. c
c c
c chkfmt1 : a long test-suite for the conversion routines. c
c c
c chkun.f : a test-suite for *some* of the routines in unary.f c
c c
c rvbr.f : Test program for Variable Block Matrix support c
c c
c c
c makefile : makefile for the sample programs chkfmt1.f, chkun.f c
c and rvbr.f produces the executables fmt.ex for format c
c (usage: make fmt.ex), un.ex for unary.f (usage: c
c make un.ex) and vbr.ex for VBR functions (usage: c
c make vbr.ex). c
c c
c----------------------------------------------------------------------c
c c
c Here are the contents of formats.f and unary.f c
c c
c----------------------------------------------------------------------c
c======================================================================c
c formats.f: c
c======================================================================c
c c
c----------------------------------------------------------------------c
c S P A R S K I T c
c----------------------------------------------------------------------c
c FORMAT CONVERSION MODULE c
c----------------------------------------------------------------------c
c contents: c
c---------- c
c csrdns : converts a row-stored sparse matrix into the dense format. c
c dnscsr : converts a dense matrix to a sparse storage format. c
c coocsr : converts coordinate to to csr format c
c coicsr : in-place conversion of coordinate to csr format c
c csrcoo : converts compressed sparse row to coordinate. c
c csrssr : converts compressed sparse row to symmetric sparse row c
c ssrcsr : converts symmetric sparse row to compressed sparse row c
c csrell : converts compressed sparse row to ellpack format c
c ellcsr : converts ellpack format to compressed sparse row format c
c csrmsr : converts compressed sparse row format to modified sparse c
c row format c
c msrcsr : converts modified sparse row format to compressed sparse c
c row format. c
c csrcsc : converts compressed sparse row format to compressed sparse c
c column format (transposition) c
c csrlnk : converts compressed sparse row to linked list format c
c lnkcsr : converts linked list format to compressed sparse row fmt c
c csrdia : converts a compressed sparse row format into a diagonal c
c format. c
c diacsr : converts a diagonal format into a compressed sparse row c
c format. c
c bsrcsr : converts a block-row sparse format into a compressed c
c sparse row format. c
c csrbsr : converts a compressed sparse row format into a block-row c
c sparse format. c
c csrbnd : converts a compressed sparse row format into a banded c
c format (linpack style). c
c bndcsr : converts a banded format (linpack style) into a compressed c
c sparse row storage. c
c csrssk : converts the compressed sparse row format to the symmetric c
c skyline format c
c sskssr : converts symmetric skyline format to symmetric sparse row c
c format. c
c csrjad : converts the csr format into the jagged diagonal format c
c jadcsr : converts the jagged-diagonal format into the csr format c
c csruss : Compressed Sparse Row to Unsymmetric Sparse Skyline format c
c usscsr : Unsymmetric Sparse Skyline format to Compressed Sparse Row c
c csrsss : Compressed Sparse Row to Symmetric Sparse Skyline format c
c ssscsr : Symmetric Sparse Skyline format to Compressed Sparse Row c
c csrvbr : Converts compressed sparse row to var block row format c
c vbrcsr : Converts var block row to compressed sparse row format c
c--------- miscalleneous additions not involving the csr format--------c
c cooell : converts coordinate to Ellpack/Itpack format c
c dcsort : sorting routine used by crsjad c
c csorted : Checks if matrix in CSR format is sorted by columns c
c----------------------------------------------------------------------c
c======================================================================c
c unary.f: c
c======================================================================c
c c
c----------------------------------------------------------------------c
c S P A R S K I T c
c----------------------------------------------------------------------c
c UNARY SUBROUTINES MODULE c
c----------------------------------------------------------------------c
c contents: c
c---------- c
c submat : extracts a submatrix from a sparse matrix. c
c filter : filters elements from a matrix according to their magnitude.c
c filterm: same as above, but for the MSR format c
c csort : sorts the elements in increasing order of columns c
c transp : in-place transposition routine (see also csrcsc in formats) c
c copmat : copy of a matrix into another matrix (both stored csr) c
c getelm : returns a(i,j) for any (i,j) from a CSR-stored matrix. c
c getdia : extracts a specified diagonal from a matrix. c
c getl : extracts lower triangular part c
c getu : extracts upper triangular part c
c levels : gets the level scheduling structure for lower triangular c
c matrices. c
c amask : extracts C = A mask M c
c rperm : permutes the rows of a matrix (B = P A) c
c cperm : permutes the columns of a matrix (B = A Q) c
c dperm : permutes both the rows and columns of a matrix (B = P A Q ) c
c dmperm : symmetric permutation of row and column (B = P A P') c
c dvperm : permutes a real vector (in-place) c
c ivperm : permutes an integer vector (in-place) c
c retmx : returns the max absolute value in each row of the matrix c
c diapos : returns the positions of the diagonal elements in A. c
c extbdg : extracts the main diagonal blocks of a matrix. c
c getbwd : returns the bandwidth information on a matrix. c
c blkfnd : finds the block-size of a matrix. c
c blkchk : checks whether a given integer is the block size of A. c
c infdia : obtains information on the diagonals of A. c
c amubdg : gets number of nonzeros in each row of A*B (as well as NNZ) c
c aplbdg : gets number of nonzeros in each row of A+B (as well as NNZ) c
c rnrms : computes the norms of the rows of A c
c cnrms : computes the norms of the columns of A c
c rscal : scales the rows of a matrix by their norms. c
c cscal : scales the columns of a matrix by their norms. c
c addblk : Adds a matrix B into a block of A. c
c get1up : Collects the first elements of each row of the upper c
c triangular portion of the matrix. c
c xtrows : extracts given rows from a matrix in CSR format. c
c csrkvstr: Finds block row partitioning of matrix in CSR format c
c csrkvstc: Finds block column partitioning of matrix in CSR format c
c kvstmerge: Merges block partitionings, for conformal row/col pattern c
c----------------------------------------------------------------------c
|
