N2G: Fits the N2G model

In AnalyzeFMRI: Functions for Analysis of fMRI Datasets Stored in the ANALYZE or NIFTI Format

Description

Fits the N2G model (1 Normal and 2 Gamma's mixture model) to a dataset using Maximum Likelihhod.

Usage

N2G(data, par.start = c(4, 2, 4, 2, 0.9, 0.05))

Arguments

data	The dataset.
par.start	The starting values for the optimization to maximize the likelihood. The parameters of the model are ordered in the vector par.start in the following way (refer to the model below) c(a, b, c, d, p1, p2)

Details

The mixture model considered is a mixture of a standard normal distribution and two Gamma functions. This model is denoted N2G.

x ~ p1 * N(0, 1) + p2 * Gamma(a, b) + (1 - p1 - p2) * -Gamma(c, d)

Value

A list with components

par	The fitted parameter values.
lims	The upper and lower thresholds for the Normal component of the fitted model

Author(s)

J. L. Marchini

See Also

N2G.Class.Probability, N2G.Likelihood.Ratio, N2G.Spatial.Mixture, N2G.Density , N2G.Likelihood , N2G.Transform, N2G.Fit , N2G.Inverse , N2G.Region

Examples

par <- c(3, 2, 3, 2, .3, .4)
data <- c(rnorm(10000), rgamma(2000, 10, 1), -rgamma(1400, 10, 1))
hist(data, n = 100, freq = FALSE)

q <- N2G.Fit(data, par, maxit = 10000, method = "BFGS")
p <- seq(-50, 50, .1)
lines(p, N2G.Density(p, q), col = 2)

AnalyzeFMRI documentation built on Oct. 5, 2021, 5:06 p.m.