lqmmControl function

Control parameters for lqmm estimation

A list of parameters for controlling the fitting process.

lqmmControl(method = "gs", LP_tol_ll = 1e-5, LP_tol_theta = 1e-5,
        check_theta = FALSE, LP_step = NULL, beta = 0.5, gamma = 1,
        reset_step = FALSE, LP_max_iter = 500, UP_tol = 1e-4,
        UP_max_iter = 20, startQR = FALSE, verbose = FALSE)

Arguments

• method: character vector that specifies the estimation method: "gs" for gradient search (default) and "df" for Nelder-Mead.
• LP_tol_ll: tolerance expressed as absolute change of the log-likelihood.
• LP_tol_theta: tolerance expressed as absolute change of theta
• check_theta: logical flag. If TRUE the algorithm performs an additional check on the change in the estimates.
• LP_step: step size (default standard deviation of response).
• beta: decreasing step factor for line search (0,1).
• gamma: nondecreasing step factor for line search (>= 1).
• reset_step: logical flag. If TRUE the step size is reset to the initial value at each iteration.
• LP_max_iter: maximum number of iterations
• UP_tol: tolerance expressed as absolute change of the scale parameter.
• UP_max_iter: maximum number of iterations.
• startQR: logical flag. If FALSE (default) the least squares estimate of the fixed effects is used as starting value of theta_x and scale. If TRUE the lqm estimate is used.
• verbose: logical flag.

Details

LP (lower loop) refers to the estimation of regression coefficients and variance-covariance parameters. UP (upper loop) refers to the estimation of the scale parameter.

Returns

a list of control parameters.

Author(s)

Marco Geraci

See Also

lqmm