theta.ml2: Theta estimation with penalty

In DavidKLim/FSCseq: Feature Selection and Clustering of RNA-seq Count Data

Description

theta.ml() with small ridge penalty to stabilize estimates, and using more efficient digamma and trigamma functions.

Usage

theta.ml2(
  y,
  mu,
  n = sum(weights),
  weights,
  t0 = 0,
  limit = 10,
  eps = .Machine$double.eps^0.25,
  trace = FALSE
)

Arguments

y	cts matrix
mu	mean parameters
n	number of samples
weights	number of weights
t0	initial estimate of dispersion scale parameter
limit	max number of iterations
eps	tolerance
trace	boolean whether to trace progress

Value

theta

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