.bhtsne {CellTrails} | R Documentation |
reduceDimensions
#' @keywords internal
#' @author Daniel C. Ellwanger
.reduceDimensions_def <- function(x, s)
CellTrails::latentSpace(x) <- CellTrails::latentSpace(x)[, seq_len(s@n)]
CellTrails::eigenvalues(x) <- CellTrails::eigenvalues(x)[seq_len(s@n)]
x
t-Distributed Stochastic Neighbor EmbeddingBarnes-Hut implementation of t-Distributed Stochastic Neighbor Embedding
.bhtsne(x, dims = 2, perplexity = 30, theta = 0.5, max_iter = 1000)
x |
A numerical matrix |
dims |
Output dimensionality |
perplexity |
Perplexity parameter (default: 30) |
theta |
Speed/accuracy trade-off (increase for less accuracy), set to 0.0 for exact tSNE (default: .5) |
max_iter |
Number of iterations (default: 1000) |
A list with the following components:
Y
Matrix containing the new representations for the objects
perplexity
See above
Daniel C. Ellwanger