TEST(Methods,HessianLocallyLinearEmbeddingSmokeTest)
{
const int N = 100;
tapkee::DenseMatrix X = swissroll(N);
kernel_callback kcb(X);
distance_callback dcb(X);
feature_vector_callback fcb(X);
tapkee::ParametersMap parameters;
parameters[tapkee::REDUCTION_METHOD] = tapkee::HESSIAN_LOCALLY_LINEAR_EMBEDDING;
parameters[tapkee::CURRENT_DIMENSION] = static_cast<tapkee::IndexType>(3);
parameters[tapkee::TARGET_DIMENSION] = static_cast<tapkee::IndexType>(2);
parameters[tapkee::NUMBER_OF_NEIGHBORS] = static_cast<tapkee::IndexType>(15);
parameters[tapkee::NEIGHBORS_METHOD] = tapkee::COVER_TREE;
parameters[tapkee::EIGEN_EMBEDDING_METHOD] = tapkee::ARPACK;
//tapkee::LoggingSingleton::instance().enable_info();
vector<int> data;
for (int i=0; i<N; i++)
data.push_back(i);
tapkee::ReturnResult result;
// should produce no error
ASSERT_NO_THROW(result = tapkee::embed(data.begin(),data.end(),kcb,dcb,fcb,parameters));
// that's normal
ASSERT_EQ(2,result.first.cols());
// that's normal
ASSERT_EQ(N,result.first.rows());
}
TapkeeOutput embedUsing(const DenseMatrix& matrix) const
{
vector<IndexType> indices(matrix.cols());
for (IndexType i=0; i<matrix.cols(); i++) indices[i] = i;
eigen_kernel_callback kcb(matrix);
eigen_distance_callback dcb(matrix);
eigen_features_callback fvcb(matrix);
return tapkee::embed(indices.begin(),indices.end(),kcb,dcb,fvcb,parameters);
}
