Rust-based t-SNE (GPU)

Performs t-SNE dimensionality reduction on the input data. This function provides a user-friendly interface with input validation before calling the Rust implementation. Leverages GPU-accelerated kNN searches. The optimisation itself still runs on the CPU (a GPU optimiser is on the roadmap).

Usage

tsne_gpu(
  data,
  knn = NULL,
  n_dim = 2L,
  perplexity = 20,
  approx_type = c("bh", "fft"),
  knn_method = c("nndescent", "exhaustive", "ivf"),
  nn_params = params_nn_gpu(),
  tsne_params = params_tsne_gpu(),
  seed = 42L,
  use_high_precision = NULL,
  .verbose = TRUE
)

Arguments

data

Numerical matrix or data frame. The data to embed of shape samples x features. Will be coerced to a matrix.

knn

Optional NearestNeighbours class. If provided, t-SNE will skip the k-nearest neighbour graph generation and use this one. Defaults to NULL. See manifoldsR::new_nearest_neighbour() for details.

n_dim

Integer. Number of dimensions in the embedding space. Currently only 2L is supported. Defaults to 2L.

perplexity

Numeric. Perplexity parameter, related to the number of nearest neighbours used in manifold learning. Typical values are between 5 and 50. Defaults to 20.0.

approx_type

Character. Approximation method for computing repulsive forces. One of "bh" for Barnes-Hut or "fft" for FFT-accelerated interpolation. Defaults to "bh". The FFT variant is only available on Unix systems.

knn_method

Character. GPU-accelerated (approximate) nearest neighbour method to use. One of "nndescent", "exhaustive", or "ivf".

nn_params

Named list. Nearest neighbour search parameters, see params_nn_gpu().

tsne_params

Named list. t-SNE (GPU) algorithm parameters, see params_tsne_gpu().

seed

Integer. Random seed for reproducibility. Defaults to 42L.

use_high_precision

Optional boolean. Gives fine-grained control over fp32 vs fp64 usage. The GPU kNN calculations will be forced into fp32.

.verbose

Logical. Controls verbosity. Defaults to TRUE.

Value

A numerical matrix with dimensions samples x n_dim containing the t-SNE embedding.

Details

The number of neighbours is derived from perplexity on the Rust side following the usual 3 * perplexity convention.