neighborhoodToArray: neighborhoodToArray

In amcdavid/HurdleNormal: 'Estimation and sampling for multi-dimensional Hurdle models on a Normal density with applications to single-cell co-expression'

neighborhoodToArray

Description

Convert neighborhood estimates fit on differing lambda paths into adjacency matrices

Usage

neighborhoodToArray(
  pathList,
  nknots,
  vnames = NULL,
  summaryFun = summarySignedL1,
  nobs,
  self_edges = FALSE
)

Arguments

pathList	a list of paths (coefficients in columns, paths in rows)
nknots	number of lambda knots. Be default, use the number of lambda in each neighborhood in pathList
vnames	a vector of names to be applied to the resulting adjacency matrix
summaryFun	function to reduce a vector-valued parameter set at each node to a scalar (representing an edge weight). Defaults to the "signed" L2 norm.
nobs	number of observations the model was fit; used to calculate BIC
self_edges	should self edges (loops) be returned in the adjacency matrix; this allows inference of intercept quantities.

Details

Join a series of paths fit node-wise into a list of adjacency matrices.

Value

list of (sparse) adjacency matrices, the number of non-zero elements (edges, before enforcing symmetry) for each matrix, the lambda for each matrix, the non-penalized, refitted pseudo log-likelihood, the number of parameters per edge, and the BIC

See Also

fitHurdle, autoGLM, interpolateEdges

amcdavid/HurdleNormal documentation built on May 14, 2022, 11:12 p.m.