#ifndef lint static char *RCSid = "$Id: standard.c,v 3.26 92/03/24 22:34:37 woo Exp Locker: woo $"; #endif /* GNUPLOT - standard.c */ /* * Copyright (C) 1986, 1987, 1990, 1991, 1992 Thomas Williams, Colin Kelley * * Permission to use, copy, and distribute this software and its * documentation for any purpose with or without fee is hereby granted, * provided that the above copyright notice appear in all copies and * that both that copyright notice and this permission notice appear * in supporting documentation. * * Permission to modify the software is granted, but not the right to * distribute the modified code. Modifications are to be distributed * as patches to released version. * * This software is provided "as is" without express or implied warranty. * * * AUTHORS * * Original Software: * Thomas Williams, Colin Kelley. * * Gnuplot 2.0 additions: * Russell Lang, Dave Kotz, John Campbell. * * Gnuplot 3.0 additions: * Gershon Elber and many others. * * Send your comments or suggestions to * info-gnuplot@ames.arc.nasa.gov. * This is a mailing list; to join it send a note to * info-gnuplot-request@ames.arc.nasa.gov. * Send bug reports to * bug-gnuplot@ames.arc.nasa.gov. */ #include #include #include "plot.h" #ifdef vms #include #else extern int errno; #endif /* vms */ extern struct value stack[STACK_DEPTH]; extern int s_p; extern double zero; struct value *pop(), *complex(), *integer(); double magnitude(), angle(), real(), imag(); /* The bessel function approximations here are from * "Computer Approximations" * by Hart, Cheney et al. * John Wiley & Sons, 1968 */ /* There appears to be a mistake in Hart, Cheney et al. on page 149. * Where it list Qn(x)/x ~ P(z*z)/Q(z*z), z = 8/x, it should read * Qn(x)/z ~ P(z*z)/Q(z*z), z = 8/x * In the functions below, Qn(x) is implementated using the later * equation. * These bessel functions are accurate to about 1e-13 */ #define PI_ON_FOUR 0.78539816339744830961566084581987572 #define PI_ON_TWO 1.57079632679489661923131269163975144 #define THREE_PI_ON_FOUR 2.35619449019234492884698253745962716 #define TWO_ON_PI 0.63661977236758134307553505349005744 static double dzero = 0.0; /* jzero for x in [0,8] * Index 5849, 19.22 digits precision */ static double pjzero[] = { 0.4933787251794133561816813446e+21, -0.11791576291076105360384408e+21, 0.6382059341072356562289432465e+19, -0.1367620353088171386865416609e+18, 0.1434354939140346111664316553e+16, -0.8085222034853793871199468171e+13, 0.2507158285536881945555156435e+11, -0.4050412371833132706360663322e+8, 0.2685786856980014981415848441e+5 }; static double qjzero[] = { 0.4933787251794133562113278438e+21, 0.5428918384092285160200195092e+19, 0.3024635616709462698627330784e+17, 0.1127756739679798507056031594e+15, 0.3123043114941213172572469442e+12, 0.669998767298223967181402866e+9, 0.1114636098462985378182402543e+7, 0.1363063652328970604442810507e+4, 0.1e+1 }; /* pzero for x in [8,inf] * Index 6548, 18.16 digits precision */ static double ppzero[] = { 0.2277909019730468430227002627e+5, 0.4134538663958076579678016384e+5, 0.2117052338086494432193395727e+5, 0.348064864432492703474453111e+4, 0.15376201909008354295771715e+3, 0.889615484242104552360748e+0 }; static double qpzero[] = { 0.2277909019730468431768423768e+5, 0.4137041249551041663989198384e+5, 0.2121535056188011573042256764e+5, 0.350287351382356082073561423e+4, 0.15711159858080893649068482e+3, 0.1e+1 }; /* qzero for x in [8,inf] * Index 6948, 18.33 digits precision */ static double pqzero[] = { -0.8922660020080009409846916e+2, -0.18591953644342993800252169e+3, -0.11183429920482737611262123e+3, -0.2230026166621419847169915e+2, -0.124410267458356384591379e+1, -0.8803330304868075181663e-2, }; static double qqzero[] = { 0.571050241285120619052476459e+4, 0.1195113154343461364695265329e+5, 0.726427801692110188369134506e+4, 0.148872312322837565816134698e+4, 0.9059376959499312585881878e+2, 0.1e+1 }; /* yzero for x in [0,8] * Index 6245, 18.78 digits precision */ static double pyzero[] = { -0.2750286678629109583701933175e+20, 0.6587473275719554925999402049e+20, -0.5247065581112764941297350814e+19, 0.1375624316399344078571335453e+18, -0.1648605817185729473122082537e+16, 0.1025520859686394284509167421e+14, -0.3436371222979040378171030138e+11, 0.5915213465686889654273830069e+8, -0.4137035497933148554125235152e+5 }; static double qyzero[] = { 0.3726458838986165881989980739e+21, 0.4192417043410839973904769661e+19, 0.2392883043499781857439356652e+17, 0.9162038034075185262489147968e+14, 0.2613065755041081249568482092e+12, 0.5795122640700729537380087915e+9, 0.1001702641288906265666651753e+7, 0.1282452772478993804176329391e+4, 0.1e+1 }; /* jone for x in [0,8] * Index 6050, 20.98 digits precision */ static double pjone[] = { 0.581199354001606143928050809e+21, -0.6672106568924916298020941484e+20, 0.2316433580634002297931815435e+19, -0.3588817569910106050743641413e+17, 0.2908795263834775409737601689e+15, -0.1322983480332126453125473247e+13, 0.3413234182301700539091292655e+10, -0.4695753530642995859767162166e+7, 0.270112271089232341485679099e+4 }; static double qjone[] = { 0.11623987080032122878585294e+22, 0.1185770712190320999837113348e+20, 0.6092061398917521746105196863e+17, 0.2081661221307607351240184229e+15, 0.5243710262167649715406728642e+12, 0.1013863514358673989967045588e+10, 0.1501793594998585505921097578e+7, 0.1606931573481487801970916749e+4, 0.1e+1 }; /* pone for x in [8,inf] * Index 6749, 18.11 digits precision */ static double ppone[] = { 0.352246649133679798341724373e+5, 0.62758845247161281269005675e+5, 0.313539631109159574238669888e+5, 0.49854832060594338434500455e+4, 0.2111529182853962382105718e+3, 0.12571716929145341558495e+1 }; static double qpone[] = { 0.352246649133679798068390431e+5, 0.626943469593560511888833731e+5, 0.312404063819041039923015703e+5, 0.4930396490181088979386097e+4, 0.2030775189134759322293574e+3, 0.1e+1 }; /* qone for x in [8,inf] * Index 7149, 18.28 digits precision */ static double pqone[] = { 0.3511751914303552822533318e+3, 0.7210391804904475039280863e+3, 0.4259873011654442389886993e+3, 0.831898957673850827325226e+2, 0.45681716295512267064405e+1, 0.3532840052740123642735e-1 }; static double qqone[] = { 0.74917374171809127714519505e+4, 0.154141773392650970499848051e+5, 0.91522317015169922705904727e+4, 0.18111867005523513506724158e+4, 0.1038187585462133728776636e+3, 0.1e+1 }; /* yone for x in [0,8] * Index 6444, 18.24 digits precision */ static double pyone[] = { -0.2923821961532962543101048748e+20, 0.7748520682186839645088094202e+19, -0.3441048063084114446185461344e+18, 0.5915160760490070618496315281e+16, -0.4863316942567175074828129117e+14, 0.2049696673745662182619800495e+12, -0.4289471968855248801821819588e+9, 0.3556924009830526056691325215e+6 }; static double qyone[] = { 0.1491311511302920350174081355e+21, 0.1818662841706134986885065935e+19, 0.113163938269888452690508283e+17, 0.4755173588888137713092774006e+14, 0.1500221699156708987166369115e+12, 0.3716660798621930285596927703e+9, 0.726914730719888456980191315e+6, 0.10726961437789255233221267e+4, 0.1e+1 }; f_real() { struct value a; push( complex(&a,real(pop(&a)), 0.0) ); } f_imag() { struct value a; push( complex(&a,imag(pop(&a)), 0.0) ); } f_arg() { struct value a; push( complex(&a,angle(pop(&a)), 0.0) ); } f_conjg() { struct value a; (void) pop(&a); push( complex(&a,real(&a),-imag(&a) )); } f_sin() { struct value a; (void) pop(&a); push( complex(&a,sin(real(&a))*cosh(imag(&a)), cos(real(&a))*sinh(imag(&a))) ); } f_cos() { struct value a; (void) pop(&a); push( complex(&a,cos(real(&a))*cosh(imag(&a)), -sin(real(&a))*sinh(imag(&a)))); } f_tan() { struct value a; register double den; (void) pop(&a); if (imag(&a) == 0.0) push( complex(&a,tan(real(&a)),0.0) ); else { den = cos(2*real(&a))+cosh(2*imag(&a)); if (den == 0.0) { undefined = TRUE; push( &a ); } else push( complex(&a,sin(2*real(&a))/den, sinh(2*imag(&a))/den) ); } } f_asin() { struct value a; register double alpha, beta, x, y; (void) pop(&a); x = real(&a); y = imag(&a); if (y == 0.0) { if (fabs(x) > 1.0) { undefined = TRUE; push(complex(&a,0.0, 0.0)); } else push( complex(&a,asin(x),0.0) ); } else { beta = sqrt((x + 1)*(x + 1) + y*y)/2 - sqrt((x - 1)*(x - 1) + y*y)/2; alpha = sqrt((x + 1)*(x + 1) + y*y)/2 + sqrt((x - 1)*(x - 1) + y*y)/2; push( complex(&a,asin(beta), log(alpha + sqrt(alpha*alpha-1))) ); } } f_acos() { struct value a; register double alpha, beta, x, y; (void) pop(&a); x = real(&a); y = imag(&a); if (y == 0.0) { if (fabs(x) > 1.0) { undefined = TRUE; push(complex(&a,0.0, 0.0)); } else push( complex(&a,acos(x),0.0) ); } else { alpha = sqrt((x + 1)*(x + 1) + y*y)/2 + sqrt((x - 1)*(x - 1) + y*y)/2; beta = sqrt((x + 1)*(x + 1) + y*y)/2 - sqrt((x - 1)*(x - 1) + y*y)/2; push( complex(&a,acos(beta), log(alpha + sqrt(alpha*alpha-1))) ); } } f_atan() { struct value a; register double x, y, u, v, w, z; (void) pop(&a); x = real(&a); y = imag(&a); if (y == 0.0) push( complex(&a,atan(x), 0.0) ); else if (x == 0.0 && fabs(y) == 1.0) { undefined = TRUE; push(complex(&a,0.0, 0.0)); } else { if (x >= 0) { u = x; v = y; } else { u = -x; v = -y; } z = atan(2*u/(1-u*u-v*v)); w = log((u*u+(v+1)*(v+1))/(u*u+(v-1)*(v-1)))/4; if (z < 0) z = z + 2*PI_ON_TWO; if (x < 0) { z = -z; w = -w; } push( complex(&a,0.5*z, w) ); } } f_sinh() { struct value a; (void) pop(&a); push( complex(&a,sinh(real(&a))*cos(imag(&a)), cosh(real(&a))*sin(imag(&a))) ); } f_cosh() { struct value a; (void) pop(&a); push( complex(&a,cosh(real(&a))*cos(imag(&a)), sinh(real(&a))*sin(imag(&a))) ); } f_tanh() { struct value a; register double den; (void) pop(&a); den = cosh(2*real(&a)) + cos(2*imag(&a)); push( complex(&a,sinh(2*real(&a))/den, sin(2*imag(&a))/den) ); } f_int() { struct value a; push( integer(&a,(int)real(pop(&a))) ); } f_abs() { struct value a; (void) pop(&a); switch (a.type) { case INT: push( integer(&a,abs(a.v.int_val)) ); break; case CMPLX: push( complex(&a,magnitude(&a), 0.0) ); } } f_sgn() { struct value a; (void) pop(&a); switch(a.type) { case INT: push( integer(&a,(a.v.int_val > 0) ? 1 : (a.v.int_val < 0) ? -1 : 0) ); break; case CMPLX: push( integer(&a,(a.v.cmplx_val.real > 0.0) ? 1 : (a.v.cmplx_val.real < 0.0) ? -1 : 0) ); break; } } f_sqrt() { struct value a; register double mag, ang; (void) pop(&a); mag = sqrt(magnitude(&a)); if (imag(&a) == 0.0 && real(&a) < 0.0) push( complex(&a,0.0,mag) ); else { if ( (ang = angle(&a)) < 0.0) ang += 2*Pi; ang /= 2; push( complex(&a,mag*cos(ang), mag*sin(ang)) ); } } f_exp() { struct value a; register double mag, ang; (void) pop(&a); mag = exp(real(&a)); ang = imag(&a); push( complex(&a,mag*cos(ang), mag*sin(ang)) ); } f_log10() { struct value a; register double l10;; (void) pop(&a); l10 = log(10.0); /***** replace with a constant! ******/ push( complex(&a,log(magnitude(&a))/l10, angle(&a)/l10) ); } f_log() { struct value a; (void) pop(&a); push( complex(&a,log(magnitude(&a)), angle(&a)) ); } f_floor() { struct value a; (void) pop(&a); switch (a.type) { case INT: push( integer(&a,(int)floor((double)a.v.int_val))); break; case CMPLX: push( integer(&a,(int)floor(a.v.cmplx_val.real))); } } f_ceil() { struct value a; (void) pop(&a); switch (a.type) { case INT: push( integer(&a,(int)ceil((double)a.v.int_val))); break; case CMPLX: push( integer(&a,(int)ceil(a.v.cmplx_val.real))); } } #ifdef GAMMA f_gamma() { extern int signgam; register double y; struct value a; y = GAMMA(real(pop(&a))); if (y > 88.0) { undefined = TRUE; push( integer(&a,0) ); } else push( complex(&a,signgam * exp(y),0.0) ); } #endif /* GAMMA */ /* bessel function approximations */ double jzero(x) double x; { double p, q, x2; int n; x2 = x * x; p = pjzero[8]; q = qjzero[8]; for (n=7; n>=0; n--) { p = p*x2 + pjzero[n]; q = q*x2 + qjzero[n]; } return(p/q); } double pzero(x) double x; { double p, q, z, z2; int n; z = 8.0 / x; z2 = z * z; p = ppzero[5]; q = qpzero[5]; for (n=4; n>=0; n--) { p = p*z2 + ppzero[n]; q = q*z2 + qpzero[n]; } return(p/q); } double qzero(x) double x; { double p, q, z, z2; int n; z = 8.0 / x; z2 = z * z; p = pqzero[5]; q = qqzero[5]; for (n=4; n>=0; n--) { p = p*z2 + pqzero[n]; q = q*z2 + qqzero[n]; } return(p/q); } double yzero(x) double x; { double p, q, x2; int n; x2 = x * x; p = pyzero[8]; q = qyzero[8]; for (n=7; n>=0; n--) { p = p*x2 + pyzero[n]; q = q*x2 + qyzero[n]; } return(p/q); } double rj0(x) double x; { if ( x <= 0.0 ) x = -x; if ( x < 8.0 ) return(jzero(x)); else return( sqrt(TWO_ON_PI/x) * (pzero(x)*cos(x-PI_ON_FOUR) - 8.0/x*qzero(x)*sin(x-PI_ON_FOUR)) ); } double ry0(x) double x; { if ( x < 0.0 ) return(dzero/dzero); /* error */ if ( x < 8.0 ) return( yzero(x) + TWO_ON_PI*rj0(x)*log(x) ); else return( sqrt(TWO_ON_PI/x) * (pzero(x)*sin(x-PI_ON_FOUR) + (8.0/x)*qzero(x)*cos(x-PI_ON_FOUR)) ); } double jone(x) double x; { double p, q, x2; int n; x2 = x * x; p = pjone[8]; q = qjone[8]; for (n=7; n>=0; n--) { p = p*x2 + pjone[n]; q = q*x2 + qjone[n]; } return(p/q); } double pone(x) double x; { double p, q, z, z2; int n; z = 8.0 / x; z2 = z * z; p = ppone[5]; q = qpone[5]; for (n=4; n>=0; n--) { p = p*z2 + ppone[n]; q = q*z2 + qpone[n]; } return(p/q); } double qone(x) double x; { double p, q, z, z2; int n; z = 8.0 / x; z2 = z * z; p = pqone[5]; q = qqone[5]; for (n=4; n>=0; n--) { p = p*z2 + pqone[n]; q = q*z2 + qqone[n]; } return(p/q); } double yone(x) double x; { double p, q, x2; int n; x2 = x * x; p = 0.0; q = qyone[8]; for (n=7; n>=0; n--) { p = p*x2 + pyone[n]; q = q*x2 + qyone[n]; } return(p/q); } double rj1(x) double x; { double v,w; v = x; if ( x < 0.0 ) x = -x; if ( x < 8.0 ) return(v*jone(x)); else { w = sqrt(TWO_ON_PI/x) * (pone(x)*cos(x-THREE_PI_ON_FOUR) - 8.0/x*qone(x)*sin(x-THREE_PI_ON_FOUR)) ; if (v < 0.0) w = -w; return( w ); } } double ry1(x) double x; { if ( x <= 0.0 ) return(dzero/dzero); /* error */ if ( x < 8.0 ) return( x*yone(x) + TWO_ON_PI*(rj1(x)*log(x) - 1.0/x) ); else return( sqrt(TWO_ON_PI/x) * (pone(x)*sin(x-THREE_PI_ON_FOUR) + (8.0/x)*qone(x)*cos(x-THREE_PI_ON_FOUR)) ); } f_besj0() { struct value a; double x; (void) pop(&a); if (imag(&a) > zero) int_error("can only do bessel functions of reals",NO_CARET); push( complex(&a,rj0(real(&a)),0.0) ); } f_besj1() { struct value a; double x; (void) pop(&a); if (imag(&a) > zero) int_error("can only do bessel functions of reals",NO_CARET); push( complex(&a,rj1(real(&a)),0.0) ); } f_besy0() { struct value a; double x; (void) pop(&a); if (imag(&a) > zero) int_error("can only do bessel functions of reals",NO_CARET); if (real(&a) > 0.0) push( complex(&a,ry0(real(&a)),0.0) ); else { push( complex(&a,0.0,0.0) ); undefined = TRUE ; } } f_besy1() { struct value a; double x; (void) pop(&a); if (imag(&a) > zero) int_error("can only do bessel functions of reals",NO_CARET); if (real(&a) > 0.0) push( complex(&a,ry1(real(&a)),0.0) ); else { push( complex(&a,0.0,0.0) ); undefined = TRUE ; } }