eigen_options Class — pytorch Architecture
Architecture documentation for the eigen_options class in SparseBlasImpl.cpp from the pytorch codebase.
Entity Profile
Source Code
aten/src/ATen/native/sparse/eigen/SparseBlasImpl.cpp lines 49–63
template <typename scalar_t, int eigen_options, typename index_t>
const Eigen::Map<Eigen::SparseMatrix<scalar_t, eigen_options, index_t>>
Tensor_to_Eigen(const at::Tensor& tensor) {
int64_t rows = tensor.size(0);
int64_t cols = tensor.size(1);
int64_t nnz = tensor._nnz();
TORCH_CHECK(tensor.values().is_contiguous(), "eigen accepts only contiguous tensor values");
auto [compressed_indices, plain_indices] = at::sparse_csr::getCompressedPlainIndices(tensor);
index_t* c_indices_ptr = compressed_indices.data_ptr<index_t>();
index_t* p_indices_ptr = plain_indices.data_ptr<index_t>();
scalar_t* values_ptr = tensor.values().data_ptr<scalar_t>();
Eigen::Map<Eigen::SparseMatrix<scalar_t, eigen_options, index_t>> map(
rows, cols, nnz, c_indices_ptr, p_indices_ptr, values_ptr);
return map;
}
Source
Analyze Your Own Codebase
Get architecture documentation, dependency graphs, and domain analysis for your codebase in minutes.
Try Supermodel Free