62 lines
1.3 KiB
C++
62 lines
1.3 KiB
C++
/* SPDX-FileCopyrightText: 2015 Blender Authors
|
|
*
|
|
* SPDX-License-Identifier: GPL-2.0-or-later */
|
|
|
|
/** \file
|
|
* \ingroup intern_eigen
|
|
*/
|
|
|
|
#ifndef __EIGEN3_SVD_C_API_CC__
|
|
#define __EIGEN3_SVD_C_API_CC__
|
|
|
|
/* Eigen gives annoying huge amount of warnings here, silence them! */
|
|
#if defined(__GNUC__) && !defined(__clang__)
|
|
# pragma GCC diagnostic ignored "-Wlogical-op"
|
|
#endif
|
|
|
|
#ifdef __EIGEN3_SVD_C_API_CC__
|
|
/* Quiet warning. */
|
|
#endif
|
|
|
|
#include <Eigen/Core>
|
|
#include <Eigen/Dense>
|
|
#include <Eigen/SVD>
|
|
|
|
#include "svd.h"
|
|
|
|
using Eigen::JacobiSVD;
|
|
|
|
using Eigen::NoQRPreconditioner;
|
|
|
|
using Eigen::ComputeThinU;
|
|
using Eigen::ComputeThinV;
|
|
|
|
using Eigen::Map;
|
|
using Eigen::MatrixXf;
|
|
using Eigen::VectorXf;
|
|
|
|
using Eigen::Matrix4f;
|
|
|
|
void EIG_svd_square_matrix(const int size, const float *matrix, float *r_U, float *r_S, float *r_V)
|
|
{
|
|
/* Since our matrix is squared, we can use thinU/V. */
|
|
unsigned int flags = (r_U ? ComputeThinU : 0) | (r_V ? ComputeThinV : 0);
|
|
|
|
/* Blender and Eigen matrices are both column-major. */
|
|
JacobiSVD<MatrixXf, NoQRPreconditioner> svd(Map<MatrixXf>((float *)matrix, size, size), flags);
|
|
|
|
if (r_U) {
|
|
Map<MatrixXf>(r_U, size, size) = svd.matrixU();
|
|
}
|
|
|
|
if (r_S) {
|
|
Map<VectorXf>(r_S, size) = svd.singularValues();
|
|
}
|
|
|
|
if (r_V) {
|
|
Map<MatrixXf>(r_V, size, size) = svd.matrixV();
|
|
}
|
|
}
|
|
|
|
#endif /* __EIGEN3_SVD_C_API_CC__ */
|