pmxNODE (version 0.1.0)

nn_eta_initializer_nm: Internal: Set initial ETA estimates in NONMEM

Description

Set the initial ETA estimates for the $OMEGA block in the NONMEM model file

Usage

nn_eta_initializer_nm(
  number,
  theta_inis,
  pop = FALSE,
  n_hidden = 5,
  eta_scale = 0.1,
  time_nn = FALSE
)

Value

Vector of initial NN parameter ETA values for one specific NN

Arguments

number

(string) Name of the NN, e.g., “1” for NN1(...)

theta_inis

(list of string) THETA initial values generated in nn_theta_initializer_nm

pop

(boolean) Whether population fit without inter-individual variability is performed (TRUE) or whether model is fitted with inter-individual variability (FALSE)

n_hidden

(numeric) Number of neurons in the hidden layer, default value is 5

eta_scale

(numeric) Initial standard deviation of random effects, default value is 0.1

time_nn

(boolean) Whether the neural network to analyze is a time-dependent neural network or not. Default values is FALSE.

Author

Dominic Bräm

Details

ETAs are fixed to 0 if

• population fit is enabled through pop=TRUE, all ETAs are fixed to 0

• the corresponding THETA is fixed to 0 due to non-activity of the neuron