File: idmax.f

package info (click to toggle)
scilab 4.1.2-6
  • links: PTS, VCS
  • area: non-free
  • in suites: lenny
  • size: 113,992 kB
  • ctags: 65,732
  • sloc: ansic: 406,468; fortran: 242,412; xml: 223,812; tcl: 46,703; sh: 10,945; ml: 9,441; makefile: 4,697; cpp: 1,354; java: 926; csh: 260; yacc: 247; perl: 130; lex: 126; asm: 72; lisp: 30
file content (43 lines) | stat: -rw-r--r-- 1,113 bytes parent folder | download | duplicates (2)
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
      integer function idmax(n, x, incx)
*
*     PURPOSE
*        finds the index of element having maximum value
*        without taking into account the nan(s)...
*
*     NOTES
*        - original version modified by Bruno (for the nan problem...)
*          (01/01/2003)
*        - this function return 1 if x has only nan components : may be
*          this is not a good behavior
*        - this function doesn't test if n<1 or incx<1 : this is done
*          by the scilab interface
*
      implicit none
      integer n, incx
      double precision x(incx,*)

      double precision xmax
      integer i, j
      external isanan
      integer  isanan

      idmax = 1

*     initialize the max with the first component being not a nan
      j = 1      
      do while ( isanan(x(1,j)) .eq. 1 )
         j = j + 1
         if ( j .gt. n ) return
      enddo
      xmax = x(1,j)
      idmax = j

*     the usual loop
      do i = j+1, n
         if ( x(1,i) .gt. xmax) then  ! a test with a nan must always return false
            xmax = x(1,i)
            idmax = i
         endif
      enddo
 
      end