??? 01/15/06 15:12 Read: times |
#107429 - here Responding to: ???'s previous message |
/* tancotf.c: Computes tan or cot of a 32-bit float as outlined in [1] Copyright (C) 2001, 2002 Jesus Calvino-Fraga, jesusc@ieee.org This library is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version. This library is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details. You should have received a copy of the GNU Lesser General Public License along with this library; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ /* [1] William James Cody and W. M. Waite. _Software manual for the elementary functions_, Englewood Cliffs, N.J.:Prentice-Hall, 1980. */ /* Version 1.0 - Initial release */ #include <math.h> #include <errno.h> #define P0 0.100000000E+1 #define P1 -0.958017723E-1 #define Q0 0.100000000E+1 #define Q1 -0.429135777E+0 #define Q2 0.971685835E-2 #define C1 1.5703125 #define C2 4.83826794897E-4 #define P(f,g) (P1*g*f+f) #define Q(g) ((Q2*g+Q1)*g+Q0) //A reasonable choice for YMAX is the integer part of B**(t/2)*PI/2: #define YMAX 6433.0 float tancotf(const float x, const int iscotan) { float f, g, xn, xnum, xden; int n; if (fabsf(x) > YMAX) { errno = ERANGE; return 0.0; } /*Round x*2*PI to the nearest integer*/ n=(x*TWO_O_PI+(x>0.0?0.5:-0.5)); /*works for +-x*/ xn=n; xnum=(int)x; xden=x-xnum; f=((xnum-xn*C1)+xden)-xn*C2; if (fabsf(f) < EPS) { xnum = f; xden = 1.0; } else { g = f*f; xnum = P(f,g); xden = Q(g); } if(n&1) //xn is odd { if(iscotan) return (-xnum/xden); else return (-xden/xnum); } else { if(iscotan) return (xden/xnum); else return (xnum/xden); } } |
Topic | Author | Date |
tangent function | 01/01/70 00:00 | |
Tangent function | 01/01/70 00:00 | |
here | 01/01/70 00:00 | |
tangent function | 01/01/70 00:00 | |
Ah, but | 01/01/70 00:00 | |
and why cannot you | 01/01/70 00:00 | |
or use SDCC | 01/01/70 00:00 | |
Yes but.. | 01/01/70 00:00 | |
CORDIC | 01/01/70 00:00 | |
Possibly useful.... | 01/01/70 00:00 | |
Yeah, but | 01/01/70 00:00 | |
tangent function | 01/01/70 00:00 | |
Requirements | 01/01/70 00:00 | |
tangent function | 01/01/70 00:00 | |
Position | 01/01/70 00:00 | |
tangent function | 01/01/70 00:00 | |
example of an answer | 01/01/70 00:00 | |
Do you understand this? | 01/01/70 00:00 | |
More Questions | 01/01/70 00:00 | |
sin/cos approximations | 01/01/70 00:00 | |
Sine etc look up tables | 01/01/70 00:00 | |
options | 01/01/70 00:00 | |
Nasty function![]() | 01/01/70 00:00 |