Trailing-Edge
-
PDP-10 Archives
-
SRI_NIC_PERM_FS_1_19910112
-
kccdist/lib/pml/dpower.c
There are 5 other files named dpower.c in the archive. Click here to see a list.
/*
* +++ NAME +++
*
* DPOWER Double precision power function
*
* +++ INDEX +++
*
* DPOWER
* machine independent routines
* math libraries
*
* +++ DESCRIPTION +++
*
* Returns double precision result of first arg raised to second
*
* +++ USAGE +++
*
* double dpower(x, y)
* double x, y;
*
* +++ REFERENCES +++
*
* Refer to Ian's brain and Harbison & Steele's CACM section 11.3.19
*
* +++ RESTRICTIONS +++
*
* For precision information refer to documentation of the
* floating point library routines called.
*
* +++ INTERNALS +++
*
* Computes DPOWER(X, Y) from:
*
* DPOWER(X, Y) = EXP(Y * DLN(X))
*
* ---
*/
/*)LIBRARY
*/
#include <stdio.h>
#include "pmluse.h"
#include "pml.h"
#define power(x, y) (dexp((y) * dln(x)))
double dpower(x, y)
double x, y;
{
extern double dexp(), dln(), dint();
if (x > 0) {
if (y > 0)
return(power(x, y));
else if (y == 0)
return(1.0);
else /* (y < 0) */
return(1.0 / power(x, y));
}
else if (x == 0) {
if (y > 0)
return(0.0);
else { /* (y <= 0) */
pmlerr(DPOW_0_TO_NON_POS);
return(MAX_NEG_DBLF);
}
}
else { /* (x < 0) */
if (y == dint(y)) {
if (y > 0)
return(-power(-x, y));
else if (y == 0)
return(1.0);
else /* (y < 0) */
return(-1.0 / power(-x, -y));
}
else {
pmlerr(DPOW_NEG_TO_FLOAT);
return(MAX_NEG_DBLF);
}
}
}