getAffinity: Calculate responsibilities.

In mnem: Mixture Nested Effects Models

Description

This function calculates the responsibilities of each component for all cells from the expected log distribution of the hidden data.

Usage

getAffinity(
  x,
  affinity = 0,
  norm = TRUE,
  logtype = 2,
  mw = NULL,
  data = matrix(0, 2, ncol(x)),
  complete = FALSE
)

Arguments

x	log odds for l cells and k components as a kxl matrix
affinity	0 for standard soft clustering, 1 for hard clustering during inference (not recommended)
norm	if TRUE normalises to probabilities (recommended)
logtype	logarithm type of the data (e.g. 2 for log2 data or exp(1) for natural)
mw	mixture weights of the components
data	data in log odds
complete	if TRUE, complete data log likelihood is considered (for very large data sets, e.g. 1000 cells and 1000 E-genes)

Value

responsibilities as a kxl matrix (k components, l cells)

Author(s)

Martin Pirkl

Examples

sim <- simData(Sgenes = 3, Egenes = 2, Nems = 2, mw = c(0.4,0.6))
data <- (sim$data - 0.5)/0.5
data <- data + rnorm(length(data), 0, 1)
result <- mnem(data, k = 2, starts = 1)
resp <- getAffinity(result$probs, mw = result$mw, data = data)

Example output

Registered S3 methods overwritten by 'RcppEigen':
  method               from         
  predict.fastLm       RcppArmadillo
  print.fastLm         RcppArmadillo
  summary.fastLm       RcppArmadillo
  print.summary.fastLm RcppArmadillo

mnem documentation built on Nov. 18, 2020, 2 a.m.