reducedDims: Perform dimensional reduction

Description Usage Arguments

View source: R/scAI_model.R

Description

Dimension reduction by PCA, t-SNE or UMAP

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
reducedDims(object, data.use = object@fit$H, do.scale = TRUE,
  do.center = TRUE, return.object = TRUE, method = "umap",
  dim.embed = 2, dim.use = NULL, perplexity = 30, theta = 0.5,
  check_duplicates = F, rand.seed = 42L, FItsne.path = NULL,
  dimPC = 40, do.fast = TRUE, weight.by.var = TRUE,
  n.neighbors = 30L, n.components = 2L, distance = "correlation",
  n.epochs = NULL, learning.rate = 1, min.dist = 0.3, spread = 1,
  set.op.mix.ratio = 1, local.connectivity = 1L,
  repulsion.strength = 1, negative.sample.rate = 5, a = NULL,
  b = NULL)

Arguments

object

scAI object

data.use

input data

do.scale

whether scale the data

do.center

whether scale and center the data

return.object

whether return scAI object

method

Method of dimensional reduction, one of tsne, FItsne and umap

dim.embed

dimensions of t-SNE embedding

dim.use

num of PCs used for t-SNE

perplexity

perplexity parameter in tsne

theta

parameter in tsne

check_duplicates

parameter in tsne

rand.seed

Set a random seed. By default, sets the seed to 42.

FItsne.path

File path of FIt-SNE

dimPC

the number of components to keep in PCA

do.fast

whether do fast PCA

weight.by.var

whether use weighted pc.scores

n.neighbors

This determines the number of neighboring points used in local approximations of manifold structure. Larger values will result in more global structure being preserved at the loss of detailed local structure. In general this parameter should often be in the range 5 to 50.

n.components

The dimension of the space to embed into.

distance

This determines the choice of metric used to measure distance in the input space.

n.epochs

the number of training epochs to be used in optimizing the low dimensional embedding. Larger values result in more accurate embeddings. If NULL is specified, a value will be selected based on the size of the input dataset (200 for large datasets, 500 for small).

learning.rate

The initial learning rate for the embedding optimization.

min.dist

This controls how tightly the embedding is allowed compress points together. Larger values ensure embedded points are moreevenly distributed, while smaller values allow the algorithm to optimise more accurately with regard to local structure. Sensible values are in the range 0.001 to 0.5.

spread

he effective scale of embedded points. In combination with min.dist this determines how clustered/clumped the embedded points are.

set.op.mix.ratio

Interpolate between (fuzzy) union and intersection as the set operation used to combine local fuzzy simplicial sets to obtain a global fuzzy simplicial sets.

local.connectivity

The local connectivity required - i.e. the number of nearest neighbors that should be assumed to be connected at a local level. The higher this value the more connected the manifold becomes locally. In practice this should be not more than the local intrinsic dimension of the manifold.

repulsion.strength

Weighting applied to negative samples in low dimensional embedding optimization. Values higher than one will result in greater weight being given to negative samples.

negative.sample.rate

The number of negative samples to select per positive sample in the optimization process. Increasing this value will result in greater repulsive force being applied, greater optimization cost, but slightly more accuracy.

a

More specific parameters controlling the embedding. If NULL, these values are set automatically as determined by min. dist and spread.

b

More specific parameters controlling the embedding. If NULL, these values are set automatically as determined by min. dist and spread.


sqjin/scAI documentation built on Nov. 19, 2020, 4:04 p.m.