Skip to main content
Ctrl+K
RAFT - Home

RAFT

  • GitHub
  • Twitter
Home
raft
cucimcudf-javacudfcugraphcumlcuprojcuspatialcuvscuxfilterdask-cudadask-cudfkvikiolibcudflibcumllibcuprojlibcuspatiallibkvikiolibrapidsmpflibrmmlibucxxnvforestraftrapids-cmakerapidsmpfrmmucxx
nightly (26.06)
nightly (26.06)stable (26.04)legacy (26.02)
RAFT - Home RAFT - Home

RAFT

  • GitHub
  • Twitter

Table of Contents

Contents:

  • Quick Start
  • Installation
  • C++ API
    • Core
      • Resources
      • logger
      • Key-Value Pair
      • NVTX
      • Interruptible
      • Operators and Functors
      • Mathematical Functions
      • Bitset
      • Bitmap
    • Linear Algebra
      • Arithmetic
      • BLAS Routines
      • Mapping and Reduction
      • Matrix Operations
      • Matrix-Vector Operations
      • Linear Algebra Solvers
    • Matrix
      • Matrix Arithmetic
      • Matrix Manipulation
      • Matrix Ordering
      • Matrix Reductions
    • Multi-dimensional Data
      • Multi-dimensional Representation
      • mdspan: Multi-dimensional Non-owning View
      • mdarray: Multi-dimensional Owning Container
      • span: One-dimensional Non-owning View
      • mdbuffer: Multi-dimensional Maybe-Owning Container
      • memory_type_dispatcher
      • temporary_device_buffer: Temporary raft::device_mdspan Producing Object
    • Multi-node Multi-GPU
    • Random
      • Data Generation
      • Univariate Random Sampling
      • Multi-Variable Random Sampling
      • Sampling Without Replacement
    • Solvers
    • Sparse
      • Sparse Types
        • COO Matrix
        • CSR Matrix
      • Sparse Linear Algebra
      • Sparse Matrix Operations
      • Sparse Solvers
    • Stats
      • Summary Statistics
      • Probability & Information Theory
      • Regression Model Scoring
      • Classification Model Scoring
      • Clustering Model Scoring
    • Utilities
  • Python API
    • Common
    • Random
    • Sparse
  • RAFT Dask API
  • Using RAFT Comms
  • Developer Guide
  • Contributing
  • C++ API
  • Sparse

Sparse#

Core to RAFT’s computational patterns for sparse data is its vocabulary of sparse types.

Contents:

  • Sparse Types
    • SparsityType
    • row_type
    • col_type
    • nnz_type
    • element_type
    • structure_view_type
    • view_type
    • structure_type
    • container_policy_type
    • container_type
    • sparse_structure()
    • sparse_structure()
    • get_nnz()
    • get_n_rows()
    • get_n_cols()
    • initialize_sparsity()
    • sparse_matrix_view()
    • structure_view()
    • get_elements()
    • sparse_matrix()
    • sparse_matrix()
    • sparse_matrix()
    • sparse_matrix()
    • ~sparse_matrix()
    • view()
    • raft::sparse_structure
    • raft::sparse_matrix_view
    • raft::sparse_matrix
    • COO Matrix
    • CSR Matrix
  • Sparse Linear Algebra
    • csr_add_calc_inds()
    • csr_add_finalize()
    • coo_degree()
    • coo_degree()
    • coo_degree_scalar()
    • coo_degree_scalar()
    • coo_degree_nz()
    • coo_degree_nz()
    • compute_graph_laplacian()
    • compute_graph_laplacian()
    • laplacian_normalized()
    • laplacian_normalized()
    • csr_row_normalize_l1()
    • csr_row_normalize_max()
    • rowNormCsr()
    • sddmm()
    • spmm()
    • coo_symmetrize()
    • coo_symmetrize()
    • symmetrize()
    • csr_transpose()
  • Sparse Matrix Operations
    • coo_remove_scalar()
    • coo_remove_scalar()
    • coo_remove_scalar()
    • coo_remove_zeros()
    • compute_duplicates_mask()
    • max_duplicates()
    • csr_row_slice_indptr()
    • csr_row_slice_populate()
    • coo_sort()
    • coo_sort()
    • coo_sort_by_weight()
  • Sparse Solvers
    • LANCZOS_WHICH
    • lanczos_compute_eigenpairs()
    • lanczos_compute_eigenpairs()
    • lanczos_compute_eigenpairs()
    • mst()
    • raft::sparse::solver::Graph_COO
    • raft::sparse::solver::lanczos_solver_config
    • raft::sparse::solver::MST_solver

previous

Solvers

next

Sparse Types

NVIDIA NVIDIA
Privacy Policy | Your Privacy Choices | Terms of Service | Accessibility | Corporate Policies | Product Security | Contact

Copyright © 2023-2026, NVIDIA Corporation.