[16269] | 1 | /*
|
---|
| 2 | * atan.h
|
---|
| 3 | * The basic idea is to exploit Pade polynomials.
|
---|
| 4 | * A lot of ideas were inspired by the cephes math library (by Stephen L. Moshier
|
---|
| 5 | * moshier@na-net.ornl.gov) as well as actual code.
|
---|
| 6 | * The Cephes library can be found here: http://www.netlib.org/cephes/
|
---|
| 7 | *
|
---|
| 8 | * Created on: Jun 23, 2012
|
---|
| 9 | * Author: Danilo Piparo, Thomas Hauth, Vincenzo Innocente
|
---|
| 10 | */
|
---|
| 11 |
|
---|
| 12 | /*
|
---|
| 13 | * VDT is free software: you can redistribute it and/or modify
|
---|
| 14 | * it under the terms of the GNU Lesser Public License as published by
|
---|
| 15 | * the Free Software Foundation, either version 3 of the License, or
|
---|
| 16 | * (at your option) any later version.
|
---|
| 17 | *
|
---|
| 18 | * This program is distributed in the hope that it will be useful,
|
---|
| 19 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
| 20 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
| 21 | * GNU Lesser Public License for more details.
|
---|
| 22 | *
|
---|
| 23 | * You should have received a copy of the GNU Lesser Public License
|
---|
| 24 | * along with this program. If not, see <http://www.gnu.org/licenses/>.
|
---|
| 25 | */
|
---|
| 26 |
|
---|
| 27 | #ifndef ATAN_H_
|
---|
| 28 | #define ATAN_H_
|
---|
| 29 |
|
---|
| 30 | #include "vdtcore_common.h"
|
---|
| 31 |
|
---|
| 32 | namespace vdt{
|
---|
| 33 |
|
---|
| 34 | namespace details{
|
---|
| 35 | const double T3PO8 = 2.41421356237309504880;
|
---|
| 36 | const double MOREBITSO2 = MOREBITS * 0.5;
|
---|
| 37 |
|
---|
| 38 | inline double get_atan_px(const double x2){
|
---|
| 39 |
|
---|
| 40 | const double PX1atan = -8.750608600031904122785E-1;
|
---|
| 41 | const double PX2atan = -1.615753718733365076637E1;
|
---|
| 42 | const double PX3atan = -7.500855792314704667340E1;
|
---|
| 43 | const double PX4atan = -1.228866684490136173410E2;
|
---|
| 44 | const double PX5atan = -6.485021904942025371773E1;
|
---|
| 45 |
|
---|
| 46 | double px = PX1atan;
|
---|
| 47 | px *= x2;
|
---|
| 48 | px += PX2atan;
|
---|
| 49 | px *= x2;
|
---|
| 50 | px += PX3atan;
|
---|
| 51 | px *= x2;
|
---|
| 52 | px += PX4atan;
|
---|
| 53 | px *= x2;
|
---|
| 54 | px += PX5atan;
|
---|
| 55 |
|
---|
| 56 | return px;
|
---|
| 57 | }
|
---|
| 58 |
|
---|
| 59 |
|
---|
| 60 | inline double get_atan_qx(const double x2){
|
---|
| 61 | const double QX1atan = 2.485846490142306297962E1;
|
---|
| 62 | const double QX2atan = 1.650270098316988542046E2;
|
---|
| 63 | const double QX3atan = 4.328810604912902668951E2;
|
---|
| 64 | const double QX4atan = 4.853903996359136964868E2;
|
---|
| 65 | const double QX5atan = 1.945506571482613964425E2;
|
---|
| 66 |
|
---|
| 67 | double qx=x2;
|
---|
| 68 | qx += QX1atan;
|
---|
| 69 | qx *=x2;
|
---|
| 70 | qx += QX2atan;
|
---|
| 71 | qx *=x2;
|
---|
| 72 | qx += QX3atan;
|
---|
| 73 | qx *=x2;
|
---|
| 74 | qx += QX4atan;
|
---|
| 75 | qx *=x2;
|
---|
| 76 | qx += QX5atan;
|
---|
| 77 |
|
---|
| 78 | return qx;
|
---|
| 79 | }
|
---|
| 80 |
|
---|
| 81 | }
|
---|
| 82 |
|
---|
| 83 |
|
---|
| 84 |
|
---|
| 85 | /// Fast Atan implementation double precision
|
---|
| 86 | inline double fast_atan(double x){
|
---|
| 87 |
|
---|
| 88 | /* make argument positive and save the sign */
|
---|
| 89 | const uint64_t sign_mask = details::getSignMask(x);
|
---|
| 90 | x=std::fabs(x);
|
---|
| 91 |
|
---|
| 92 | /* range reduction */
|
---|
| 93 | const double originalx=x;
|
---|
| 94 |
|
---|
| 95 | double y = details::PIO4;
|
---|
| 96 | double factor = details::MOREBITSO2;
|
---|
| 97 | x = (x-1.0) / (x+1.0);
|
---|
| 98 |
|
---|
| 99 | if( originalx > details::T3PO8 ) {
|
---|
| 100 | y = details::PIO2;
|
---|
| 101 | factor = details::MOREBITS;
|
---|
| 102 | x = -1.0 / originalx ;
|
---|
| 103 | }
|
---|
| 104 | if ( originalx <= 0.66 ) {
|
---|
| 105 | y = 0.;
|
---|
| 106 | factor = 0.;
|
---|
| 107 | x = originalx;
|
---|
| 108 | }
|
---|
| 109 |
|
---|
| 110 | const double x2 = x * x;
|
---|
| 111 |
|
---|
| 112 | const double px = details::get_atan_px(x2);
|
---|
| 113 | const double qx = details::get_atan_qx(x2);
|
---|
| 114 |
|
---|
| 115 | //double res = y +x * x2 * px / qx + x +factor;
|
---|
| 116 |
|
---|
| 117 | const double poq=px / qx;
|
---|
| 118 |
|
---|
| 119 | double res = x * x2 * poq + x;
|
---|
| 120 | res+=y;
|
---|
| 121 |
|
---|
| 122 | res+=factor;
|
---|
| 123 |
|
---|
| 124 | return details::dpORuint64(res,sign_mask);
|
---|
| 125 | }
|
---|
| 126 |
|
---|
| 127 | //------------------------------------------------------------------------------
|
---|
| 128 | /// Fast Atan implementation single precision
|
---|
| 129 | inline float fast_atanf( float xx ) {
|
---|
| 130 |
|
---|
| 131 | const uint32_t sign_mask = details::getSignMask(xx);
|
---|
| 132 |
|
---|
| 133 | float x= std::fabs(xx);
|
---|
| 134 | const float x0=x;
|
---|
| 135 | float y=0.0f;
|
---|
| 136 |
|
---|
| 137 | /* range reduction */
|
---|
| 138 | if( x0 > 0.4142135623730950f ){ // * tan pi/8
|
---|
| 139 | x = (x0-1.0f)/(x0+1.0f);
|
---|
| 140 | y = details::PIO4F;
|
---|
| 141 | }
|
---|
| 142 | if( x0 > 2.414213562373095f ){ // tan 3pi/8
|
---|
| 143 | x = -( 1.0f/x0 );
|
---|
| 144 | y = details::PIO2F;
|
---|
| 145 | }
|
---|
| 146 |
|
---|
| 147 |
|
---|
| 148 | const float x2 = x * x;
|
---|
| 149 | y +=
|
---|
| 150 | ((( 8.05374449538e-2f * x2
|
---|
| 151 | - 1.38776856032E-1f) * x2
|
---|
| 152 | + 1.99777106478E-1f) * x2
|
---|
| 153 | - 3.33329491539E-1f) * x2 * x
|
---|
| 154 | + x;
|
---|
| 155 |
|
---|
| 156 | return details::spORuint32(y,sign_mask);
|
---|
| 157 | }
|
---|
| 158 |
|
---|
| 159 | //------------------------------------------------------------------------------
|
---|
| 160 |
|
---|
| 161 | }// end of vdt
|
---|
| 162 |
|
---|
| 163 | #endif // end of atan
|
---|