1DTFTRI(1)                LAPACK routine (version 3.2)                DTFTRI(1)
2
3
4

NAME

6       DTFTRI  -  computes  the inverse of a triangular matrix A stored in RFP
7       format
8

SYNOPSIS

10       SUBROUTINE DTFTRI( TRANSR, UPLO, DIAG, N, A, INFO )
11
12           CHARACTER      TRANSR, UPLO, DIAG
13
14           INTEGER        INFO, N
15
16           DOUBLE         PRECISION A( 0: * )
17

PURPOSE

19       DTFTRI computes the inverse of a triangular matrix A stored in RFP for‐
20       mat.  This is a Level 3 BLAS version of the algorithm.
21

ARGUMENTS

23       TRANSR    (input) CHARACTER
24                 = 'N':  The Normal TRANSR of RFP A is stored;
25                 = 'T':  The Transpose TRANSR of RFP A is stored.
26
27       UPLO    (input) CHARACTER
28               = 'U':  A is upper triangular;
29               = 'L':  A is lower triangular.
30
31       DIAG    (input) CHARACTER
32               = 'N':  A is non-unit triangular;
33               = 'U':  A is unit triangular.
34
35       N       (input) INTEGER
36               The order of the matrix A.  N >= 0.
37
38       A       (input/output) DOUBLE PRECISION  array, dimension (0:nt-1);
39               nt=N*(N+1)/2.  On  entry,  the triangular factor of a Hermitian
40               Positive Definite  matrix  A  in  RFP  format.  RFP  format  is
41               described by TRANSR, UPLO, and N as follows: If TRANSR = 'N'
42               then RFP A is (0:N,0:k-1) when N is even; k=N/2. RFP A is
43               (0:N-1,0:k)  when  N is odd; k=N/2. IF TRANSR = 'T' then RFP is
44               the transpose of RFP A as defined when TRANSR = 'N'.  The  con‐
45               tents  of  RFP  A are defined by UPLO as follows: If UPLO = 'U'
46               the RFP A contains the nt elements of upper packed A; If UPLO =
47               'L'  the  RFP A contains the nt elements of lower packed A. The
48               LDA of RFP A is (N+1)/2 when TRANSR = 'T'. When TRANSR  is  'N'
49               the  LDA is N+1 when N is even and N is odd. See the Note below
50               for more details.  On exit, the  (triangular)  inverse  of  the
51               original matrix, in the same storage format.
52
53       INFO    (output) INTEGER
54               = 0: successful exit
55               < 0: if INFO = -i, the i-th argument had an illegal value
56               >  0:  if  INFO  =  i,  A(i,i) is exactly zero.  The triangular
57               matrix is singular and its inverse can not be computed.
58

FURTHER DETAILS

60       We first consider Rectangular Full Packed (RFP) Format when N is  even.
61       We give an example where N = 6.
62           AP is Upper             AP is Lower
63        00 01 02 03 04 05       00
64           11 12 13 14 15       10 11
65              22 23 24 25       20 21 22
66                 33 34 35       30 31 32 33
67                    44 45       40 41 42 43 44
68                       55       50 51 52 53 54 55
69       Let TRANSR = 'N'. RFP holds AP as follows:
70       For  UPLO  =  'U'  the  upper trapezoid A(0:5,0:2) consists of the last
71       three columns of AP upper. The lower triangle  A(4:6,0:2)  consists  of
72       the transpose of the first three columns of AP upper.
73       For  UPLO  =  'L'  the lower trapezoid A(1:6,0:2) consists of the first
74       three columns of AP lower. The upper triangle  A(0:2,0:2)  consists  of
75       the transpose of the last three columns of AP lower.
76       This covers the case N even and TRANSR = 'N'.
77              RFP A                   RFP A
78             03 04 05                33 43 53
79             13 14 15                00 44 54
80             23 24 25                10 11 55
81             33 34 35                20 21 22
82             00 44 45                30 31 32
83             01 11 55                40 41 42
84             02 12 22                50 51 52
85       Now let TRANSR = 'T'. RFP A in both UPLO cases is just the transpose of
86       RFP A above. One therefore gets:
87                RFP A                   RFP A
88          03 13 23 33 00 01 02    33 00 10 20 30 40 50
89          04 14 24 34 44 11 12    43 44 11 21 31 41 51
90          05 15 25 35 45 55 22    53 54 55 22 32 42 52
91       We first consider Rectangular Full Packed (RFP) Format when N  is  odd.
92       We give an example where N = 5.
93          AP is Upper                 AP is Lower
94        00 01 02 03 04              00
95           11 12 13 14              10 11
96              22 23 24              20 21 22
97                 33 34              30 31 32 33
98                    44              40 41 42 43 44
99       Let TRANSR = 'N'. RFP holds AP as follows:
100       For  UPLO  =  'U'  the  upper trapezoid A(0:4,0:2) consists of the last
101       three columns of AP upper. The lower triangle  A(3:4,0:1)  consists  of
102       the transpose of the first two columns of AP upper.
103       For  UPLO  =  'L'  the lower trapezoid A(0:4,0:2) consists of the first
104       three columns of AP lower. The upper triangle  A(0:1,1:2)  consists  of
105       the transpose of the last two columns of AP lower.
106       This covers the case N odd and TRANSR = 'N'.
107              RFP A                   RFP A
108             02 03 04                00 33 43
109             12 13 14                10 11 44
110             22 23 24                20 21 22
111             00 33 34                30 31 32
112             01 11 44                40 41 42
113       Now let TRANSR = 'T'. RFP A in both UPLO cases is just the transpose of
114       RFP A above. One therefore gets:
115                RFP A                   RFP A
116          02 12 22 00 01             00 10 20 30 40 50
117          03 13 23 33 11             33 11 21 31 41 51
118          04 14 24 34 44             43 44 22 32 42 52
119
120
121
122 LAPACK routine (version 3.2)    November 2008                       DTFTRI(1)
Impressum