LogisticRegression: Clustering projection using logistic regression from the...

In elolab/ILoReg: ILoReg: a tool for high-resolution cell population identification from scRNA-Seq data

Clustering projection using logistic regression from the LiblineaR R package

Description

The function implements a script that downsamples data a dataset, trains a logistic regression classifier model and then projects its clustering onto itself using a trained L1-regularized logistic regression model.

Usage

LogisticRegression(
  training.sparse.matrix = NULL,
  training.ident = NULL,
  C = 0.3,
  reg.type = "L1",
  test.sparse.matrix = NULL,
  d = 0.3
)

Arguments

training.sparse.matrix	A sparse matrix (dgCMatrix) containing training sample's gene expression data with genes in rows and cells in columns. Default is NULL.
training.ident	A named factor containing sample's cluster labels for each cell in training.sparse.matrix. Default is NULL.
C	Cost of constraints violation in L1-regularized logistic regression (C). Default is 0.3.
reg.type	"L1" for LASSO and "L2" for Ridge. Default is "L1".
test.sparse.matrix	A sparse matrix (dgCMatrix) containing test sample's gene expression data with genes in rows and cells in columns. Default is NULL.
d	A numeric smaller than 1 and greater than 0 that determines how many cells per cluster should be down- and oversampled (d in N/k*d), where N is the total number of cells and k the number of clusters. Default is 0.3.

Value

a list containing the output of the LiblineaR prediction

elolab/ILoReg documentation built on March 28, 2022, 1:17 a.m.