rliger (version 1.0.0)

suggestLambda: Visually suggest appropriate lambda value

Description

Can be used to select appropriate value of lambda for factorization of particular dataset. Plot alignment and agreement for various test values of lambda. Most appropriate lambda is likely around the "elbow" of the alignment plot (when alignment stops increasing). This will likely also correspond to slower decrease in agreement. Depending on number of cores used, this process can take 10-20 minutes.

Usage

suggestLambda(
  object,
  k,
  lambda.test = NULL,
  rand.seed = 1,
  num.cores = 1,
  thresh = 1e-04,
  max.iters = 100,
  knn_k = 20,
  k2 = 500,
  ref_dataset = NULL,
  resolution = 1,
  gen.new = FALSE,
  nrep = 1,
  return.data = FALSE,
  return.raw = FALSE,
  verbose = TRUE
)

Arguments

object

liger object. Should normalize, select genes, and scale before calling.

k

Number of factors to use in test factorizations. See optimizeALS documentation.

lambda.test

Vector of lambda values to test. If not given, use default set spanning 0.25 to 60

rand.seed

Random seed for reproducibility (default 1).

num.cores

Number of cores to use for optimizing factorizations in parallel (default 1).

thresh

Convergence threshold. Convergence occurs when |obj0-obj|/(mean(obj0,obj)) < thresh

max.iters

Maximum number of block coordinate descent iterations to perform

knn_k

Number of nearest neighbors for within-dataset knn in quantileAlignSNF (default 20).

k2

Horizon parameter for quantileAlignSNF (default 500).

ref_dataset

Reference dataset for quantileAlignSNF (defaults to larger dataset).

resolution

Resolution for quantileAlignSNF (default 1).

gen.new

Do not use optimizeNewLambda in factorizations. Recommended to set TRUE when looking at only a small range of lambdas (ie. 1:7) (default FALSE)

nrep

Number restarts to perform at each lambda value tested (increase to produce smoother curve if results unclear) (default 1).

return.data

Whether to return list of data matrices (raw) or dataframe (processed) instead of ggplot object (default FALSE).

return.raw

If return.results TRUE, whether to return raw data (in format described below), or dataframe used to produce ggplot object. Raw data is matrix of alignment values for each lambda value tested (each column represents a different rep for nrep).(default FALSE)

verbose

Print progress bar/messages (TRUE by default)

Value

Matrix of results if indicated or ggplot object. Plots alignment vs. lambda to console.

Examples

# NOT RUN {
# Requires preprocessed liger object
# examine plot for most appropriate lambda, use multiple cores for faster results
suggestLambda(ligerex, k = 20, num.cores = 4)
# }