This function attempts to fix starting values by comparing the analytic gradient to a numerically approximated gradient. Parameters with a difference between the analytic and numeric gradient that exceeds 'maxdiff' will be reduced by a factor of 'reduce' in each iteration until the average absolute difference between analytic and numeric gradients is lower than 'tol'. Only off-diagonal elements in omega, sigma, kappa, lowertri or rho matrices or any element in beta matrices are adjusted.
fixstart(x, reduce = 0.5, maxdiff = 0.1, tol = 0.01, maxtry = 25)
A 'psychonetrics' model
The factor with which problematic parameters are reduced in each iteration.
Maximum difference between analytic and numeric gradient to be considered problematic.
Average absolute difference between analytic and numeric gradient that is considered acceptable.
Maximum number of iterations to attempt to fix starting values.
Sacha Epskamp