get_LVs: Calculates the latent variables using the estimators for each...

Description Usage Arguments Details Value

View source: R/get_LVs.R

Description

This is an internal function used for the iterative estimation of the Latent Variables.

Usage

1
get_LVs(nodes, max_iterations, loggers, convergence_threshold = 1e-04)

Arguments

nodes

A listenv of objects of the R6Class Node which are initialised. The nodes also need to be connected.

max_iterations

An integer indicating the maximum number of iterations before execution of LV estimation is halted when the convergence criterion is not met beforehand.

loggers

A listenv of R6Class based loggers and/or reporters. All loggers or reports must implement a log_status() method. Implemented loggers or reporters are ComponentLogger, IterationReporter, DurationLogger, and ConvergenceLogger.

convergence_threshold

A double indicating the maximum error before the iterations are assumed to have converged. It is compared to the difference between the latent variables of the current and previous iteration. If this difference is less, the algorithm is considered to have converged. A convergence threshold of a difference of 0.0001 between the Sum of Squared Errors of two subsequent iterations is used as a default.

Details

A convergence threshold of a difference of 0.0001 between the Sum of Squared Errors of two subsequent iterations is used as a default.

Value

A listenv of connected, initialized, and estimated nodes


GeertPostma/pathmodelr documentation built on Oct. 5, 2021, 4:17 p.m.