2
UMFPACK V5.1.0 (May 31, 2007), Control:
3
Matrix entry defined as: double
4
Int (generic integer) defined as: UF_long
7
1: dense row parameter: 0.2
8
"dense" rows have > max (16, (0.2)*16*sqrt(n_col) entries)
9
2: dense column parameter: 0.2
10
"dense" columns have > max (16, (0.2)*16*sqrt(n_row) entries)
11
3: pivot tolerance: 0.1
12
4: block size for dense matrix kernels: 32
14
6: initial allocation ratio: 0.7
15
7: max iterative refinement steps: 2
16
12: 2-by-2 pivot tolerance: 0.01
17
13: Q fixed during numerical factorization: 0 (auto)
18
14: AMD dense row/col parameter: 10
19
"dense" rows/columns have > max (16, (10)*sqrt(n)) entries
20
Only used if the AMD ordering is used.
21
15: diagonal pivot tolerance: 0.001
22
Only used if diagonal pivoting is attempted.
23
16: scaling: 1 (divide each row by sum of abs. values in each row)
24
17: frontal matrix allocation ratio: 0.5
26
19: AMD and COLAMD aggressive absorption: 1 (yes)
28
The following options can only be changed at compile-time:
29
8: BLAS library used: Fortran BLAS. size of BLAS integer: 4
30
9: compiled for ANSI C
31
10: CPU timer is POSIX times ( ) routine.
32
11: compiled for normal operation (debugging disabled)
33
computer/operating system: Linux
34
size of int: 4 UF_long: 8 Int: 8 pointer: 8 double: 8 Entry: 8 (in bytes)
39
estimates (upper bound) for numeric LU:
41
memory needed: 0.08 (MB)
45
numeric factorization:
48
actual numeric LU statistics:
50
memory needed: 0.06 (MB)
54
UMFPACK V5.1.0 (May 31, 2007), Info:
55
matrix entry defined as: double
56
Int (generic integer) defined as: UF_long
57
BLAS library used: Fortran BLAS. size of BLAS integer: 4
59
CPU timer: POSIX times ( ) routine.
60
number of rows in matrix A: 67
61
number of columns in matrix A: 67
62
entries in matrix A: 294
63
memory usage reported in: 16-byte Units
65
size of UF_long: 8 bytes
66
size of pointer: 8 bytes
67
size of numerical entry: 8 bytes
69
strategy used: unsymmetric
70
ordering used: colamd on A
71
modify Q during factorization: yes
72
prefer diagonal pivoting: no
73
pivots with zero Markowitz cost: 1
74
submatrix S after removing zero-cost pivots:
75
number of "dense" rows: 0
76
number of "dense" columns: 0
77
number of empty rows: 0
78
number of empty columns 0
79
submatrix S not square or diagonal not preserved
80
symbolic factorization defragmentations: 1
81
symbolic memory usage (Units): 1595
82
symbolic memory usage (MBytes): 0.0
83
Symbolic size (Units): 241
84
Symbolic size (MBytes): 0
85
symbolic factorization CPU time (sec): 0.00
86
symbolic factorization wallclock time(sec): 0.00
88
matrix scaled: yes (divided each row by sum of abs values in each row)
89
minimum sum (abs (rows of A)): 1.00000e+00
90
maximum sum (abs (rows of A)): 6.59006e+00
92
symbolic/numeric factorization: upper bound actual %
93
variable-sized part of Numeric object:
94
initial size (Units) 1517 1448 95%
95
peak size (Units) 3917 2423 62%
96
final size (Units) 981 435 44%
97
Numeric final size (Units) 1381 802 58%
98
Numeric final size (MBytes) 0.0 0.0 58%
99
peak memory usage (Units) 5133 3639 71%
100
peak memory usage (MBytes) 0.1 0.1 71%
101
numeric factorization flops 1.41920e+04 2.50100e+03 18%
102
nz in L (incl diagonal) 542 323 60%
103
nz in U (incl diagonal) 902 339 38%
104
nz in L+U (incl diagonal) 1377 595 43%
105
largest front (# entries) 483 80 17%
106
largest # rows in front 21 10 48%
107
largest # columns in front 23 11 48%
109
initial allocation ratio used: 0.7
110
# of forced updates due to frontal growth: 0
111
nz in L (incl diagonal), if none dropped 323
112
nz in U (incl diagonal), if none dropped 339
113
number of small entries dropped 0
114
nonzeros on diagonal of U: 67
115
min abs. value on diagonal of U: 2.74e-02
116
max abs. value on diagonal of U: 2.28e+00
117
estimate of reciprocal of condition number: 1.20e-02
118
indices in compressed pattern: 249
119
numerical values stored in Numeric object: 605
120
numeric factorization defragmentations: 1
121
numeric factorization reallocations: 1
122
costly numeric factorization reallocations: 0
123
numeric factorization CPU time (sec): 0.00
124
numeric factorization wallclock time (sec): 0.00
126
solve flops: 1.19000e+03
127
iterative refinement steps taken: 0
128
iterative refinement steps attempted: 0
129
solve CPU time (sec): 0.00
130
solve wall clock time (sec): 0.00
132
total symbolic + numeric + solve flops: 3.69100e+03
134
norm (A*x-b): 3.10862447E-15
135
norm (A*x-b): 2.13162821E-14
136
norm (A*x-b): 2.13162821E-14