Eigen::SparseVector<double> normProbVector(Eigen::SparseVector<double> P_vec)
{
VectorXd P_dense_vec = (VectorXd)P_vec;
Eigen::SparseVector<double> P_norm;
if (P_dense_vec == VectorXd::Zero(P_vec.size()))
{
P_norm = P_vec;
}
else{
double P_positive = 0; double P_negative = 0;
for (int row_idx = 0; row_idx < P_vec.size(); row_idx++){
P_positive = (P_vec.coeff(row_idx) > 0) ? (P_positive + P_vec.coeff(row_idx)) : P_positive;
P_negative = (P_vec.coeff(row_idx) > 0) ? P_negative : (P_negative + P_vec.coeff(row_idx));
}
if (fabs(P_positive) < fabs(P_negative)){
P_norm = -P_vec / fabs(P_negative);
}
else{
P_norm = P_vec / fabs(P_positive);
}
for (int row_idx = 0; row_idx < P_vec.size(); row_idx++){
P_norm.coeffRef(row_idx) = (P_norm.coeff(row_idx)<0) ? 0 : P_norm.coeff(row_idx);
}
}
P_norm.prune(TOLERANCE);
return P_norm;
}
Eigen::SparseVector<double> pinv_vector(Eigen::SparseVector<double> pinvvec)
{
Eigen::SparseVector<double> singularValues_inv;
singularValues_inv.resize(pinvvec.size());
for (int i = 0; i<pinvvec.size(); ++i) {
singularValues_inv.coeffRef(i) = (fabs(pinvvec.coeff(i)) > TOLERANCE) ? 1.0 / pinvvec.coeff(i) : 0;
}
singularValues_inv.prune(TOLERANCE);
return singularValues_inv;
}
IGL_INLINE void igl::diag(
const Eigen::SparseVector<T>& V,
Eigen::SparseMatrix<T>& X)
{
// clear and resize output
Eigen::DynamicSparseMatrix<T, Eigen::RowMajor> dyn_X(V.size(),V.size());
dyn_X.reserve(V.size());
// loop over non-zeros
for(typename Eigen::SparseVector<T>::InnerIterator it(V); it; ++it)
{
dyn_X.coeffRef(it.index(),it.index()) += it.value();
}
X = Eigen::SparseMatrix<T>(dyn_X);
}