blob: 4f5ad6f239612f2832019abfbbb507ddc562d854 (
plain) (
blame)
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
|
# Copyright(c) 1986 Association of Universities for Research in Astronomy Inc.
# ARCZ -- Vector reciprocal with checking for zero divisors. If the result
# of a divide would be undefined a user supplied function is called to get the
# output pixel value.
#
# NOTE: in the interests of simplicity a somewhat arbitrary tolerance is used
# to check for an undefined divide, i.e., a divide by zero or a divide by a
# number small enough to cause floating point overflow. A better way to do
# this would be to provide a machine dependent version of this operator in
# host$as which catches the hardware exception rather than using a comparison.
procedure arczd (a, b, c, npix, errfcn)
double a # numerator
double b[ARB], c[ARB] # divisor, and output arrays
int npix # number of pixels
double errfcn() # user function, called on divide by zero
int i
double divisor
double tol
extern errfcn()
errchk errfcn
begin
if (a == 0.0D0) {
call aclrd (c, npix)
return
}
tol = 1.0D-20
do i = 1, npix {
divisor = b[i]
# The following is most efficient when the data tends to be
# positive.
if (divisor < tol)
if (divisor > -tol) {
c[i] = errfcn (a)
next
}
c[i] = a / divisor
}
end
|
