read_gpt: Read GPT Object from File

In danheck/gpt: Generalized Processing Tree Models

Read GPT Object from File

Description

Returns a gpt object and checks model file and parameter restrictions

Usage

read_gpt(file, latent, restrictions = NULL)

Arguments

file	path to model file
latent	type of latent continuous distribution(s) (e.g., "normal", "gamma", "exgauss", etc.). See gpt_fit for details
restrictions	list of restrictions for MPT and continuous parameters, e.g., list("g=.5", "do=dn", "mu1=mu2=mu3")

Details

The model file should have one line per MPT branch, with entries separated by semicolons in the following order:

Tree; Category; MPT equations; Continuous Parameters 1; Cont. Parameters 2

In the model file, the latent distributions are specified by parameters corresponding to the distribution of each processing branch (separated by commas). The definition of the parameters is listed in gpt_fit. An example for latent=c("normal","normal") is:

#tree	; cat	; mpt	; normal 1
	; normal 2
tree1	; hit	; do	; m1_d,s1	; m2_d,s2
tree1	; hit	; (1-do)*g	; m1_g,s1	; m2_g,s2
tree1	; miss	; (1-do)*(1-g)	; m1_g,s1	; m2_g,s2
tree2	; fa	; (1-dn)*g	; m1_g,s1	; m2_g,s2
tree2	; cr	; (1-dn)*(1-g)	; m1_g,s1	; m2_g,s2
tree2	; cr	; dn	; m1_d,s1	; m2_d,s2

Examples

file <- paste0(path.package("gpt"), "/models/2htm_2normal.txt")
read_gpt(file=file, latent=c("normal", "normal"))
# with restrictions:
read_gpt(file=file, latent=c("normal", "normal"),
         restrictions=list("g=.5","s1=s2=1", "m1_g=m1_d"))
             
# list of example model files:
list.files(path = paste0(path.package("gpt"),"/models"))

danheck/gpt documentation built on Feb. 12, 2024, 6:21 a.m.