EmbedSWNE: Embeds NMF factors and samples in 2 dimensions

Description Usage Arguments Value

View source: R/swne_plotting.R

Description

Embeds NMF factors and samples in 2 dimensions

Usage

1
2
3
4
EmbedSWNE(H, SNN = NULL, alpha.exp = 1, snn.exp = 1, n_pull = NULL,
  proj.method = "umap", pca.red = F, dist.use = "cosine",
  snn.factor.proj = T, n.neighbors = max(round(nrow(H)/3), 4),
  min.dist = 0.5)

Arguments

H

NMF factors (factors x samples)

SNN

Shared nearest neighbors matrix (or other similarity matrix)

alpha.exp

Increasing alpha.exp increases how much the NMF factors "pull" the samples

snn.exp

Decreasing snn.exp increases the effect of the similarity matrix on the embedding

n_pull

Number of factors pulling on each sample. Must be >= 3

proj.method

Method to use for projecting factors. Currently only supports "sammon"

pca.red

Whether or not to run PCA on transposed H matrix before calculating factor distances

dist.use

Similarity function to use for calculating factor positions. Options include pearson (correlation), IC (mutual information), cosine, euclidean.

n.neighbors

Number of neighbors for UMAP factor projection

min.dist

Minimum distance for UMAP factor projection

Value

A list of factor (H.coords) and sample coordinates (sample.coords) in 2D


yanwu2014/swne documentation built on Dec. 14, 2018, 3:07 p.m.