Welcome to mirror list, hosted at ThFree Co, Russian Federation.

git.blender.org/blender.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
path: root/extern/ceres/internal/ceres/dogleg_strategy.cc
diff options
context:
space:
mode:
Diffstat (limited to 'extern/ceres/internal/ceres/dogleg_strategy.cc')
-rw-r--r--extern/ceres/internal/ceres/dogleg_strategy.cc60
1 files changed, 27 insertions, 33 deletions
diff --git a/extern/ceres/internal/ceres/dogleg_strategy.cc b/extern/ceres/internal/ceres/dogleg_strategy.cc
index ecc6b882338..03ae22f7e57 100644
--- a/extern/ceres/internal/ceres/dogleg_strategy.cc
+++ b/extern/ceres/internal/ceres/dogleg_strategy.cc
@@ -49,7 +49,7 @@ namespace internal {
namespace {
const double kMaxMu = 1.0;
const double kMinMu = 1e-8;
-}
+} // namespace
DoglegStrategy::DoglegStrategy(const TrustRegionStrategy::Options& options)
: linear_solver_(options.linear_solver),
@@ -122,8 +122,8 @@ TrustRegionStrategy::Summary DoglegStrategy::ComputeStep(
//
jacobian->SquaredColumnNorm(diagonal_.data());
for (int i = 0; i < n; ++i) {
- diagonal_[i] = std::min(std::max(diagonal_[i], min_diagonal_),
- max_diagonal_);
+ diagonal_[i] =
+ std::min(std::max(diagonal_[i], min_diagonal_), max_diagonal_);
}
diagonal_ = diagonal_.array().sqrt();
@@ -171,9 +171,8 @@ TrustRegionStrategy::Summary DoglegStrategy::ComputeStep(
// The gradient, the Gauss-Newton step, the Cauchy point,
// and all calculations involving the Jacobian have to
// be adjusted accordingly.
-void DoglegStrategy::ComputeGradient(
- SparseMatrix* jacobian,
- const double* residuals) {
+void DoglegStrategy::ComputeGradient(SparseMatrix* jacobian,
+ const double* residuals) {
gradient_.setZero();
jacobian->LeftMultiply(residuals, gradient_.data());
gradient_.array() /= diagonal_.array();
@@ -187,8 +186,7 @@ void DoglegStrategy::ComputeCauchyPoint(SparseMatrix* jacobian) {
Jg.setZero();
// The Jacobian is scaled implicitly by computing J * (D^-1 * (D^-1 * g))
// instead of (J * D^-1) * (D^-1 * g).
- Vector scaled_gradient =
- (gradient_.array() / diagonal_.array()).matrix();
+ Vector scaled_gradient = (gradient_.array() / diagonal_.array()).matrix();
jacobian->RightMultiply(scaled_gradient.data(), Jg.data());
alpha_ = gradient_.squaredNorm() / Jg.squaredNorm();
}
@@ -217,7 +215,7 @@ void DoglegStrategy::ComputeTraditionalDoglegStep(double* dogleg) {
// Case 2. The Cauchy point and the Gauss-Newton steps lie outside
// the trust region. Rescale the Cauchy point to the trust region
// and return.
- if (gradient_norm * alpha_ >= radius_) {
+ if (gradient_norm * alpha_ >= radius_) {
dogleg_step = -(radius_ / gradient_norm) * gradient_;
dogleg_step_norm_ = radius_;
dogleg_step.array() /= diagonal_.array();
@@ -242,14 +240,12 @@ void DoglegStrategy::ComputeTraditionalDoglegStep(double* dogleg) {
// = alpha * -gradient' gauss_newton_step - alpha^2 |gradient|^2
const double c = b_dot_a - a_squared_norm;
const double d = sqrt(c * c + b_minus_a_squared_norm *
- (pow(radius_, 2.0) - a_squared_norm));
-
- double beta =
- (c <= 0)
- ? (d - c) / b_minus_a_squared_norm
- : (radius_ * radius_ - a_squared_norm) / (d + c);
- dogleg_step = (-alpha_ * (1.0 - beta)) * gradient_
- + beta * gauss_newton_step_;
+ (pow(radius_, 2.0) - a_squared_norm));
+
+ double beta = (c <= 0) ? (d - c) / b_minus_a_squared_norm
+ : (radius_ * radius_ - a_squared_norm) / (d + c);
+ dogleg_step =
+ (-alpha_ * (1.0 - beta)) * gradient_ + beta * gauss_newton_step_;
dogleg_step_norm_ = dogleg_step.norm();
dogleg_step.array() /= diagonal_.array();
VLOG(3) << "Dogleg step size: " << dogleg_step_norm_
@@ -345,13 +341,13 @@ void DoglegStrategy::ComputeSubspaceDoglegStep(double* dogleg) {
// correctly determined.
const double kCosineThreshold = 0.99;
const Vector2d grad_minimum = subspace_B_ * minimum + subspace_g_;
- const double cosine_angle = -minimum.dot(grad_minimum) /
- (minimum.norm() * grad_minimum.norm());
+ const double cosine_angle =
+ -minimum.dot(grad_minimum) / (minimum.norm() * grad_minimum.norm());
if (cosine_angle < kCosineThreshold) {
LOG(WARNING) << "First order optimality seems to be violated "
<< "in the subspace method!\n"
- << "Cosine of angle between x and B x + g is "
- << cosine_angle << ".\n"
+ << "Cosine of angle between x and B x + g is " << cosine_angle
+ << ".\n"
<< "Taking a regular dogleg step instead.\n"
<< "Please consider filing a bug report if this "
<< "happens frequently or consistently.\n";
@@ -423,15 +419,17 @@ Vector DoglegStrategy::MakePolynomialForBoundaryConstrainedProblem() const {
const double trB = subspace_B_.trace();
const double r2 = radius_ * radius_;
Matrix2d B_adj;
+ // clang-format off
B_adj << subspace_B_(1, 1) , -subspace_B_(0, 1),
- -subspace_B_(1, 0) , subspace_B_(0, 0);
+ -subspace_B_(1, 0) , subspace_B_(0, 0);
+ // clang-format on
Vector polynomial(5);
polynomial(0) = r2;
polynomial(1) = 2.0 * r2 * trB;
polynomial(2) = r2 * (trB * trB + 2.0 * detB) - subspace_g_.squaredNorm();
- polynomial(3) = -2.0 * (subspace_g_.transpose() * B_adj * subspace_g_
- - r2 * detB * trB);
+ polynomial(3) =
+ -2.0 * (subspace_g_.transpose() * B_adj * subspace_g_ - r2 * detB * trB);
polynomial(4) = r2 * detB * detB - (B_adj * subspace_g_).squaredNorm();
return polynomial;
@@ -565,10 +563,8 @@ LinearSolver::Summary DoglegStrategy::ComputeGaussNewtonStep(
// of Jx = -r and later set x = -y to avoid having to modify
// either jacobian or residuals.
InvalidateArray(n, gauss_newton_step_.data());
- linear_solver_summary = linear_solver_->Solve(jacobian,
- residuals,
- solve_options,
- gauss_newton_step_.data());
+ linear_solver_summary = linear_solver_->Solve(
+ jacobian, residuals, solve_options, gauss_newton_step_.data());
if (per_solve_options.dump_format_type == CONSOLE ||
(per_solve_options.dump_format_type != CONSOLE &&
@@ -641,9 +637,7 @@ void DoglegStrategy::StepIsInvalid() {
reuse_ = false;
}
-double DoglegStrategy::Radius() const {
- return radius_;
-}
+double DoglegStrategy::Radius() const { return radius_; }
bool DoglegStrategy::ComputeSubspaceModel(SparseMatrix* jacobian) {
// Compute an orthogonal basis for the subspace using QR decomposition.
@@ -701,8 +695,8 @@ bool DoglegStrategy::ComputeSubspaceModel(SparseMatrix* jacobian) {
subspace_g_ = subspace_basis_.transpose() * gradient_;
- Eigen::Matrix<double, 2, Eigen::Dynamic, Eigen::RowMajor>
- Jb(2, jacobian->num_rows());
+ Eigen::Matrix<double, 2, Eigen::Dynamic, Eigen::RowMajor> Jb(
+ 2, jacobian->num_rows());
Jb.setZero();
Vector tmp;
generated by cgit v1.2.3 (git 2.25.1) at 2024-08-31 17:16:43 +0300