Discursive space with UMAP given a distance matrix

Construct a 2-dimensional "discursive space" embedding given a distance matrix.

Usage

# S3 method for matrix
umap(dist_mx, include_data = FALSE, df = TRUE, ...)

Arguments

dist_mx

Distance matrix

include_data

By default, to save space the data (distance matrix) is not returned

df

Return a tibble with columns document, x, and y (default) or the output of umap.

...

Other parameters passed to umap::umap()

Value

Object of class umap, with components layout (coordinates of items), knn (k-nearest neighbors matrices), config (UMAP configuration) or tidied dataframe, per argument df

Examples

gamma = rdirichlet(26, 1, 5)
rownames(gamma) = letters
h_gamma = hellinger(gamma)
embedded = umap(h_gamma, df = TRUE, verbose = TRUE)
#> [2023-11-15 08:24:50.33582]  starting umap
#> [2023-11-15 08:24:50.369645]  creating graph of nearest neighbors
#> [2023-11-15 08:24:50.388917]  creating initial embedding
#> [2023-11-15 08:24:50.398499]  optimizing embedding
#> [2023-11-15 08:24:50.513047]  done