| 1 | < | /* | 
| 2 | < | * Copyright (C) 2000-2004  Object Oriented Parallel Simulation Engine (OOPSE) project | 
| 3 | < | * | 
| 4 | < | * Contact: oopse@oopse.org | 
| 5 | < | * | 
| 6 | < | * This program is free software; you can redistribute it and/or | 
| 7 | < | * modify it under the terms of the GNU Lesser General Public License | 
| 8 | < | * as published by the Free Software Foundation; either version 2.1 | 
| 9 | < | * of the License, or (at your option) any later version. | 
| 10 | < | * All we ask is that proper credit is given for our work, which includes | 
| 11 | < | * - but is not limited to - adding the above copyright notice to the beginning | 
| 12 | < | * of your source code files, and to any copyright notice that you may distribute | 
| 13 | < | * with programs based on this work. | 
| 14 | < | * | 
| 15 | < | * This program is distributed in the hope that it will be useful, | 
| 16 | < | * but WITHOUT ANY WARRANTY; without even the implied warranty of | 
| 17 | < | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the | 
| 18 | < | * GNU Lesser General Public License for more details. | 
| 19 | < | * | 
| 20 | < | * You should have received a copy of the GNU Lesser General Public License | 
| 21 | < | * along with this program; if not, write to the Free Software | 
| 22 | < | * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA. | 
| 1 | > | /* | 
| 2 | > | * Copyright (c) 2005 The University of Notre Dame. All Rights Reserved. | 
| 3 |  | * | 
| 4 | + | * The University of Notre Dame grants you ("Licensee") a | 
| 5 | + | * non-exclusive, royalty free, license to use, modify and | 
| 6 | + | * redistribute this software in source and binary code form, provided | 
| 7 | + | * that the following conditions are met: | 
| 8 | + | * | 
| 9 | + | * 1. Acknowledgement of the program authors must be made in any | 
| 10 | + | *    publication of scientific results based in part on use of the | 
| 11 | + | *    program.  An acceptable form of acknowledgement is citation of | 
| 12 | + | *    the article in which the program was described (Matthew | 
| 13 | + | *    A. Meineke, Charles F. Vardeman II, Teng Lin, Christopher | 
| 14 | + | *    J. Fennell and J. Daniel Gezelter, "OOPSE: An Object-Oriented | 
| 15 | + | *    Parallel Simulation Engine for Molecular Dynamics," | 
| 16 | + | *    J. Comput. Chem. 26, pp. 252-271 (2005)) | 
| 17 | + | * | 
| 18 | + | * 2. Redistributions of source code must retain the above copyright | 
| 19 | + | *    notice, this list of conditions and the following disclaimer. | 
| 20 | + | * | 
| 21 | + | * 3. Redistributions in binary form must reproduce the above copyright | 
| 22 | + | *    notice, this list of conditions and the following disclaimer in the | 
| 23 | + | *    documentation and/or other materials provided with the | 
| 24 | + | *    distribution. | 
| 25 | + | * | 
| 26 | + | * This software is provided "AS IS," without a warranty of any | 
| 27 | + | * kind. All express or implied conditions, representations and | 
| 28 | + | * warranties, including any implied warranty of merchantability, | 
| 29 | + | * fitness for a particular purpose or non-infringement, are hereby | 
| 30 | + | * excluded.  The University of Notre Dame and its licensors shall not | 
| 31 | + | * be liable for any damages suffered by licensee as a result of | 
| 32 | + | * using, modifying or distributing the software or its | 
| 33 | + | * derivatives. In no event will the University of Notre Dame or its | 
| 34 | + | * licensors be liable for any lost revenue, profit or data, or for | 
| 35 | + | * direct, indirect, special, consequential, incidental or punitive | 
| 36 | + | * damages, however caused and regardless of the theory of liability, | 
| 37 | + | * arising out of the use of or inability to use software, even if the | 
| 38 | + | * University of Notre Dame has been advised of the possibility of | 
| 39 | + | * such damages. | 
| 40 |  | */ | 
| 41 | < |  | 
| 41 | > |  | 
| 42 |  | /** | 
| 43 |  | * @file SquareMatrix3.hpp | 
| 44 |  | * @author Teng Lin | 
| 77 |  | /** copy  constructor */ | 
| 78 |  | SquareMatrix3(const SquareMatrix<Real, 3>& m)  : SquareMatrix<Real, 3>(m) { | 
| 79 |  | } | 
| 80 | < |  | 
| 80 | > |  | 
| 81 |  | SquareMatrix3( const Vector3<Real>& eulerAngles) { | 
| 82 |  | setupRotMat(eulerAngles); | 
| 83 |  | } | 
| 103 |  | return *this; | 
| 104 |  | } | 
| 105 |  |  | 
| 106 | + |  | 
| 107 | + | SquareMatrix3<Real>& operator =(const Quaternion<Real>& q) { | 
| 108 | + | this->setupRotMat(q); | 
| 109 | + | return *this; | 
| 110 | + | } | 
| 111 | + |  | 
| 112 |  | /** | 
| 113 |  | * Sets this matrix to a rotation matrix by three euler angles | 
| 114 |  | * @ param euler | 
| 134 |  | ctheta = cos(theta); | 
| 135 |  | cpsi = cos(psi); | 
| 136 |  |  | 
| 137 | < | data_[0][0] = cpsi * cphi - ctheta * sphi * spsi; | 
| 138 | < | data_[0][1] = cpsi * sphi + ctheta * cphi * spsi; | 
| 139 | < | data_[0][2] = spsi * stheta; | 
| 137 | > | this->data_[0][0] = cpsi * cphi - ctheta * sphi * spsi; | 
| 138 | > | this->data_[0][1] = cpsi * sphi + ctheta * cphi * spsi; | 
| 139 | > | this->data_[0][2] = spsi * stheta; | 
| 140 |  |  | 
| 141 | < | data_[1][0] = -spsi * ctheta - ctheta * sphi * cpsi; | 
| 142 | < | data_[1][1] = -spsi * stheta + ctheta * cphi * cpsi; | 
| 143 | < | data_[1][2] = cpsi * stheta; | 
| 141 | > | this->data_[1][0] = -spsi * ctheta - ctheta * sphi * cpsi; | 
| 142 | > | this->data_[1][1] = -spsi * stheta + ctheta * cphi * cpsi; | 
| 143 | > | this->data_[1][2] = cpsi * stheta; | 
| 144 |  |  | 
| 145 | < | data_[2][0] = stheta * sphi; | 
| 146 | < | data_[2][1] = -stheta * cphi; | 
| 147 | < | data_[2][2] = ctheta; | 
| 145 | > | this->data_[2][0] = stheta * sphi; | 
| 146 | > | this->data_[2][1] = -stheta * cphi; | 
| 147 | > | this->data_[2][2] = ctheta; | 
| 148 |  | } | 
| 149 |  |  | 
| 150 |  |  | 
| 177 |  | Quaternion<Real> q; | 
| 178 |  | Real t, s; | 
| 179 |  | Real ad1, ad2, ad3; | 
| 180 | < | t = data_[0][0] + data_[1][1] + data_[2][2] + 1.0; | 
| 180 | > | t = this->data_[0][0] + this->data_[1][1] + this->data_[2][2] + 1.0; | 
| 181 |  |  | 
| 182 |  | if( t > 0.0 ){ | 
| 183 |  |  | 
| 184 |  | s = 0.5 / sqrt( t ); | 
| 185 |  | q[0] = 0.25 / s; | 
| 186 | < | q[1] = (data_[1][2] - data_[2][1]) * s; | 
| 187 | < | q[2] = (data_[2][0] - data_[0][2]) * s; | 
| 188 | < | q[3] = (data_[0][1] - data_[1][0]) * s; | 
| 186 | > | q[1] = (this->data_[1][2] - this->data_[2][1]) * s; | 
| 187 | > | q[2] = (this->data_[2][0] - this->data_[0][2]) * s; | 
| 188 | > | q[3] = (this->data_[0][1] - this->data_[1][0]) * s; | 
| 189 |  | } else { | 
| 190 |  |  | 
| 191 | < | ad1 = fabs( data_[0][0] ); | 
| 192 | < | ad2 = fabs( data_[1][1] ); | 
| 193 | < | ad3 = fabs( data_[2][2] ); | 
| 191 | > | ad1 = fabs( this->data_[0][0] ); | 
| 192 | > | ad2 = fabs( this->data_[1][1] ); | 
| 193 | > | ad3 = fabs( this->data_[2][2] ); | 
| 194 |  |  | 
| 195 |  | if( ad1 >= ad2 && ad1 >= ad3 ){ | 
| 196 |  |  | 
| 197 | < | s = 2.0 * sqrt( 1.0 + data_[0][0] - data_[1][1] - data_[2][2] ); | 
| 198 | < | q[0] = (data_[1][2] + data_[2][1]) / s; | 
| 197 | > | s = 2.0 * sqrt( 1.0 + this->data_[0][0] - this->data_[1][1] - this->data_[2][2] ); | 
| 198 | > | q[0] = (this->data_[1][2] + this->data_[2][1]) / s; | 
| 199 |  | q[1] = 0.5 / s; | 
| 200 | < | q[2] = (data_[0][1] + data_[1][0]) / s; | 
| 201 | < | q[3] = (data_[0][2] + data_[2][0]) / s; | 
| 200 | > | q[2] = (this->data_[0][1] + this->data_[1][0]) / s; | 
| 201 | > | q[3] = (this->data_[0][2] + this->data_[2][0]) / s; | 
| 202 |  | } else if ( ad2 >= ad1 && ad2 >= ad3 ) { | 
| 203 | < | s = sqrt( 1.0 + data_[1][1] - data_[0][0] - data_[2][2] ) * 2.0; | 
| 204 | < | q[0] = (data_[0][2] + data_[2][0]) / s; | 
| 205 | < | q[1] = (data_[0][1] + data_[1][0]) / s; | 
| 203 | > | s = sqrt( 1.0 + this->data_[1][1] - this->data_[0][0] - this->data_[2][2] ) * 2.0; | 
| 204 | > | q[0] = (this->data_[0][2] + this->data_[2][0]) / s; | 
| 205 | > | q[1] = (this->data_[0][1] + this->data_[1][0]) / s; | 
| 206 |  | q[2] = 0.5 / s; | 
| 207 | < | q[3] = (data_[1][2] + data_[2][1]) / s; | 
| 207 | > | q[3] = (this->data_[1][2] + this->data_[2][1]) / s; | 
| 208 |  | } else { | 
| 209 |  |  | 
| 210 | < | s = sqrt( 1.0 + data_[2][2] - data_[0][0] - data_[1][1] ) * 2.0; | 
| 211 | < | q[0] = (data_[0][1] + data_[1][0]) / s; | 
| 212 | < | q[1] = (data_[0][2] + data_[2][0]) / s; | 
| 213 | < | q[2] = (data_[1][2] + data_[2][1]) / s; | 
| 210 | > | s = sqrt( 1.0 + this->data_[2][2] - this->data_[0][0] - this->data_[1][1] ) * 2.0; | 
| 211 | > | q[0] = (this->data_[0][1] + this->data_[1][0]) / s; | 
| 212 | > | q[1] = (this->data_[0][2] + this->data_[2][0]) / s; | 
| 213 | > | q[2] = (this->data_[1][2] + this->data_[2][1]) / s; | 
| 214 |  | q[3] = 0.5 / s; | 
| 215 |  | } | 
| 216 |  | } | 
| 231 |  | */ | 
| 232 |  | Vector3<Real> toEulerAngles() { | 
| 233 |  | Vector3<Real> myEuler; | 
| 234 | < | Real phi,theta,psi,eps; | 
| 235 | < | Real ctheta,stheta; | 
| 234 | > | Real phi; | 
| 235 | > | Real theta; | 
| 236 | > | Real psi; | 
| 237 | > | Real ctheta; | 
| 238 | > | Real stheta; | 
| 239 |  |  | 
| 240 |  | // set the tolerance for Euler angles and rotation elements | 
| 241 |  |  | 
| 242 | < | theta = acos(std::min(1.0, std::max(-1.0,data_[2][2]))); | 
| 243 | < | ctheta = data_[2][2]; | 
| 242 | > | theta = acos(std::min(1.0, std::max(-1.0,this->data_[2][2]))); | 
| 243 | > | ctheta = this->data_[2][2]; | 
| 244 |  | stheta = sqrt(1.0 - ctheta * ctheta); | 
| 245 |  |  | 
| 246 |  | // when sin(theta) is close to 0, we need to consider singularity | 
| 253 |  |  | 
| 254 |  | if (fabs(stheta) <= oopse::epsilon){ | 
| 255 |  | psi = 0.0; | 
| 256 | < | phi = atan2(-data_[1][0], data_[0][0]); | 
| 256 | > | phi = atan2(-this->data_[1][0], this->data_[0][0]); | 
| 257 |  | } | 
| 258 |  | // we only have one unique solution | 
| 259 |  | else{ | 
| 260 | < | phi = atan2(data_[2][0], -data_[2][1]); | 
| 261 | < | psi = atan2(data_[0][2], data_[1][2]); | 
| 260 | > | phi = atan2(this->data_[2][0], -this->data_[2][1]); | 
| 261 | > | psi = atan2(this->data_[0][2], this->data_[1][2]); | 
| 262 |  | } | 
| 263 |  |  | 
| 264 |  | //wrap phi and psi, make sure they are in the range from 0 to 2*Pi | 
| 279 |  | Real determinant() const { | 
| 280 |  | Real x,y,z; | 
| 281 |  |  | 
| 282 | < | x = data_[0][0] * (data_[1][1] * data_[2][2] - data_[1][2] * data_[2][1]); | 
| 283 | < | y = data_[0][1] * (data_[1][2] * data_[2][0] - data_[1][0] * data_[2][2]); | 
| 284 | < | z = data_[0][2] * (data_[1][0] * data_[2][1] - data_[1][1] * data_[2][0]); | 
| 282 | > | x = this->data_[0][0] * (this->data_[1][1] * this->data_[2][2] - this->data_[1][2] * this->data_[2][1]); | 
| 283 | > | y = this->data_[0][1] * (this->data_[1][2] * this->data_[2][0] - this->data_[1][0] * this->data_[2][2]); | 
| 284 | > | z = this->data_[0][2] * (this->data_[1][0] * this->data_[2][1] - this->data_[1][1] * this->data_[2][0]); | 
| 285 |  |  | 
| 286 |  | return(x + y + z); | 
| 287 |  | } | 
| 288 | + |  | 
| 289 | + | /** Returns the trace of this matrix. */ | 
| 290 | + | Real trace() const { | 
| 291 | + | return this->data_[0][0] + this->data_[1][1] + this->data_[2][2]; | 
| 292 | + | } | 
| 293 |  |  | 
| 294 |  | /** | 
| 295 |  | * Sets the value of this matrix to  the inversion of itself. | 
| 296 |  | * @note since simple algorithm can be applied to inverse the 3 by 3 matrix, we hide the | 
| 297 |  | * implementation of inverse in SquareMatrix class | 
| 298 |  | */ | 
| 299 | < | SquareMatrix3<Real>  inverse() { | 
| 299 | > | SquareMatrix3<Real>  inverse() const { | 
| 300 |  | SquareMatrix3<Real> m; | 
| 301 |  | double det = determinant(); | 
| 302 |  | if (fabs(det) <= oopse::epsilon) { | 
| 304 |  | //"This is a runtime or a programming error in your application."); | 
| 305 |  | } | 
| 306 |  |  | 
| 307 | < | m(0, 0) = data_[1][1]*data_[2][2] - data_[1][2]*data_[2][1]; | 
| 308 | < | m(1, 0) = data_[1][2]*data_[2][0] - data_[1][0]*data_[2][2]; | 
| 309 | < | m(2, 0) = data_[1][0]*data_[2][1] - data_[1][1]*data_[2][0]; | 
| 310 | < | m(0, 1) = data_[2][1]*data_[0][2] - data_[2][2]*data_[0][1]; | 
| 311 | < | m(1, 1) = data_[2][2]*data_[0][0] - data_[2][0]*data_[0][2]; | 
| 312 | < | m(2, 1) = data_[2][0]*data_[0][1] - data_[2][1]*data_[0][0]; | 
| 313 | < | m(0, 2) = data_[0][1]*data_[1][2] - data_[0][2]*data_[1][1]; | 
| 314 | < | m(1, 2) = data_[0][2]*data_[1][0] - data_[0][0]*data_[1][2]; | 
| 315 | < | m(2, 2) = data_[0][0]*data_[1][1] - data_[0][1]*data_[1][0]; | 
| 307 | > | m(0, 0) = this->data_[1][1]*this->data_[2][2] - this->data_[1][2]*this->data_[2][1]; | 
| 308 | > | m(1, 0) = this->data_[1][2]*this->data_[2][0] - this->data_[1][0]*this->data_[2][2]; | 
| 309 | > | m(2, 0) = this->data_[1][0]*this->data_[2][1] - this->data_[1][1]*this->data_[2][0]; | 
| 310 | > | m(0, 1) = this->data_[2][1]*this->data_[0][2] - this->data_[2][2]*this->data_[0][1]; | 
| 311 | > | m(1, 1) = this->data_[2][2]*this->data_[0][0] - this->data_[2][0]*this->data_[0][2]; | 
| 312 | > | m(2, 1) = this->data_[2][0]*this->data_[0][1] - this->data_[2][1]*this->data_[0][0]; | 
| 313 | > | m(0, 2) = this->data_[0][1]*this->data_[1][2] - this->data_[0][2]*this->data_[1][1]; | 
| 314 | > | m(1, 2) = this->data_[0][2]*this->data_[1][0] - this->data_[0][0]*this->data_[1][2]; | 
| 315 | > | m(2, 2) = this->data_[0][0]*this->data_[1][1] - this->data_[0][1]*this->data_[1][0]; | 
| 316 |  |  | 
| 317 |  | m /= det; | 
| 318 |  | return m; | 
| 466 |  | v = v.transpose(); | 
| 467 |  | return ; | 
| 468 |  | } | 
| 469 | + |  | 
| 470 | + | /** | 
| 471 | + | * Return the multiplication of two matrixes  (m1 * m2). | 
| 472 | + | * @return the multiplication of two matrixes | 
| 473 | + | * @param m1 the first matrix | 
| 474 | + | * @param m2 the second matrix | 
| 475 | + | */ | 
| 476 | + | template<typename Real> | 
| 477 | + | inline SquareMatrix3<Real> operator *(const SquareMatrix3<Real>& m1, const SquareMatrix3<Real>& m2) { | 
| 478 | + | SquareMatrix3<Real> result; | 
| 479 | + |  | 
| 480 | + | for (unsigned int i = 0; i < 3; i++) | 
| 481 | + | for (unsigned int j = 0; j < 3; j++) | 
| 482 | + | for (unsigned int k = 0; k < 3; k++) | 
| 483 | + | result(i, j)  += m1(i, k) * m2(k, j); | 
| 484 | + |  | 
| 485 | + | return result; | 
| 486 | + | } | 
| 487 | + |  | 
| 488 | + | template<typename Real> | 
| 489 | + | inline SquareMatrix3<Real> outProduct(const Vector3<Real>& v1, const Vector3<Real>& v2) { | 
| 490 | + | SquareMatrix3<Real> result; | 
| 491 | + |  | 
| 492 | + | for (unsigned int i = 0; i < 3; i++) { | 
| 493 | + | for (unsigned int j = 0; j < 3; j++) { | 
| 494 | + | result(i, j)  = v1[i] * v2[j]; | 
| 495 | + | } | 
| 496 | + | } | 
| 497 | + |  | 
| 498 | + | return result; | 
| 499 | + | } | 
| 500 | + |  | 
| 501 | + |  | 
| 502 |  | typedef SquareMatrix3<double> Mat3x3d; | 
| 503 |  | typedef SquareMatrix3<double> RotMat3x3d; | 
| 504 |  |  |