influence.combi: Evaluate the influence function

In combi: Compositional omics model based visual integration

Description

Evaluate the influence function

Usage

## S3 method for class 'combi'
influence(modelObj, samples = is.null(View), Dim = 1,
  View = NULL)

Arguments

modelObj	The model object
samples	A boolean, should we look at sample variables? Throws an error otherwise
Dim, View	Integers, the dimension and views required

Details

Especially the influence of the different views on the latent variable or gradient estimation may be of interest. The influence values are not all calculated. Rather, the score values and inverse jacobian are returned so they can easily be calculated

Value

A list with components

score	The evaluation of the score function
InvJac	The inverted jacobian matrix

combi documentation built on Nov. 8, 2020, 5:34 p.m.