setup_tSNE: Setup tSNE config.

Description

Setup tSNE config.

Usage

setup_tSNE(
  k = 2L,
  initial_dims = 50L,
  perplexity = 30,
  theta = 0.5,
  check_duplicates = TRUE,
  pca = TRUE,
  partial_pca = FALSE,
  max_iter = 1000L,
  verbose = getOption("verbose", FALSE),
  is_distance = FALSE,
  Y_init = NULL,
  pca_center = TRUE,
  pca_scale = FALSE,
  normalize = TRUE,
  stop_lying_iter = if (is.null(Y_init)) 250L else 0L,
  mom_switch_iter = if (is.null(Y_init)) 250L else 0L,
  momentum = 0.5,
  final_momentum = 0.8,
  eta = 200,
  exaggeration_factor = 12,
  num_threads = NULL,
  features = NULL
)

Value

tSNEConfig object.

Arguments

k: Integer [1, Inf): Number of components.
initial_dims: Integer [1, Inf): Initial dimensions.
perplexity: Numeric [0, Inf): Perplexity.
theta: Numeric [0, 1]: Speed/accuracy trade-off.
check_duplicates: Logical: If TRUE, check for duplicates.
pca: Logical: If TRUE, perform PCA.
partial_pca: Logical: If TRUE, perform partial PCA.
max_iter: Integer [1, Inf): Maximum number of iterations.
verbose: Logical: If TRUE, print messages.
is_distance: Logical: If TRUE, x is a distance matrix.
Y_init: Optional Matrix: Initial Y matrix.
pca_center: Logical: If TRUE, center PCA.
pca_scale: Logical: If TRUE, scale PCA.
normalize: Logical: If TRUE, normalize.
stop_lying_iter: Integer [0, Inf): Stop lying iterations.
mom_switch_iter: Integer [0, Inf): Momentum switch iterations.
momentum: Numeric: Momentum.
final_momentum: Numeric: Final momentum.
eta: Numeric: Eta.
exaggeration_factor: Numeric: Exaggeration factor.
num_threads: Optional Integer: Deprecated and ignored. Threads come from the execution config: set n_workers_algorithm in the execution_config passed to decomp.
features: Optional Character vector: Names of at least 2 distinct feature columns to decompose. NULL decomposes all numeric features.

Author

EDG

Details

Get more information on the config by running ?Rtsne::Rtsne.

Examples

Run this code

tSNE_config <- setup_tSNE(k = 3L)
tSNE_config

Run the code above in your browser using DataLab