FSCseq_predict: Prediction via FSCseq

Description Usage Arguments Value Author(s) References

View source: R/FSCseq.R

Description

Performs prediction of cluster membership via trained model

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
FSCseq_predict(
  X = NULL,
  fit,
  cts_train = NULL,
  cts_pred,
  SF_train = NULL,
  SF_pred,
  maxit = 100,
  eps = 1e-04
)

Arguments

X

(optional) design matrix of dimension n by p

fit

FSCseq results object

cts_train

Training count data matrix of dimension g by n_train

cts_pred

Prediction/test count data matrix of dimension g by n_pred

SF_train

Vector of length n_train, size factors for training subjects. Can be accessed from the fit.

SF_pred

Vector of length n_pred, size factors for prediction subjects

maxit

Maximum number of iterations to run if prediction batches are estimated, default is 100.

eps

Tolerance for relative change in Q function convergence criterion if prediction batches are estimated, default is 1e-4.

Value

list containing outputs final_clusters: vector of length n of resulting clusters, wts: k by n matrix of E step weights

Author(s)

David K. Lim, deelim@live.unc.edu

References

https://github.com/DavidKLim/FSCseq


DavidKLim/FSCseq documentation built on Dec. 12, 2021, 3:46 a.m.