dNBllcol: A score function for the estimation of the column scores in...

In RCM: Fit row-column association models with the negative binomial distribution for the microbiome

Description

A score function for the estimation of the column scores in an unconstrained RC(M) model

Usage

dNBllcol(
  beta,
  X,
  reg,
  thetas,
  muMarg,
  k,
  p,
  n,
  colWeights,
  nLambda,
  cMatK,
  allowMissingness,
  naId,
  ...
)

Arguments

beta	vector of length p+1+1+(k-1): p row scores, 1 centering, one normalization and (k-1) orhtogonality lagrangian multipliers
X	the nxp data matrix
reg	a nx1 regressor matrix: outer product of rowScores and psis
thetas	nxp matrix with the dispersion parameters (converted to matrix for numeric reasons)
muMarg	the nxp offset
k	an integer, the dimension of the RC solution
p	an integer, the number of taxa
n	an integer, the number of samples
colWeights	the weights used for the restrictions
nLambda	an integer, the number of restrictions
cMatK	the lower dimensions of the colScores
allowMissingness	A boolean, are missing values present
naId	The numeric index of the missing values in X
...	further arguments passed on to the jacobian

Value

A vector of length p+1+1+(k-1) with evaluations of the derivative of lagrangian

RCM documentation built on Nov. 8, 2020, 5:22 p.m.