getFit: Dimensionality reduction for visualization

In montesmariana/semcloud: Post-processing of token-level clouds

Dimensionality reduction for visualization

Description

Wrapper around metaMDS and Rtsne for dimensionality reduction with certain set parameters.

Usage

getFit(d, dim = 2, technique, perp = 30, seed = 8541)

Arguments

d	distance matrix as a matrix
dim	number of dimensions, by default 2. As argument k for metaMDS and dims for Rtsne
technique	either "mds" to run metaMDS, "tsne" to run Rtsne or "umap" to run umap
perp	perplexity value for Rtsne, default is 30. This value is ignored when metaMDS is run.
seed	seed to keep randomness at check

Details

metaMDS is run with trymax=20 and trace=FALSE by default, so that the search for best solution is not printed. Rtsne uses as default parameters theta=0.0, check.duplicates = FALSE, max_iter = 1000.

Value

output of either metaMDS or Rtsne.

montesmariana/semcloud documentation built on April 15, 2022, 6:57 a.m.