Default parameters for CAGRA-style kNN search

Usage

params_sc_cagra(
  k_query = 15L,
  ann_dist = "cosine",
  k = NULL,
  k_build = NULL,
  refine_sweeps = 0L,
  max_iters = NULL,
  n_trees = NULL,
  delta = 0.001,
  rho = NULL,
  beam_width = NULL,
  max_beam_iters = NULL,
  n_entry_points = NULL
)

Arguments

k_query: Integer. Number of neighbours to identify.
ann_dist: Character. Distance metric to use. One of "euclidean" or "cosine".
k: Optional integer. Final node degree of the CAGRA navigational graph. If NULL, defaults to 30 on the Rust side.
k_build: Optional integer. Number of k-neighbours during the NNDescent build phase before CAGRA pruning. If NULL, defaults to 1.5 * k on the Rust side.
refine_sweeps: Integer. Number of refinement sweeps during graph generation.
max_iters: Optional integer. Maximum iterations for the NNDescent rounds. If NULL, determined automatically.
n_trees: Optional integer. Number of trees to use in the initial GPU-accelerated forest. If NULL, determined automatically.
delta: Numeric. Termination criterion for the NNDescent iterations.
rho: Optional numeric. Sampling rate during NNDescent iterations. If NULL, determined automatically.
beam_width: Optional integer. Beam width during querying. If NULL, determined automatically.
max_beam_iters: Optional integer. Maximum beam iterations. If NULL, determined automatically.
n_entry_points: Optional integer. Number of entry points into the graph. If NULL, determined automatically.

Value

A list with the parameters.