Rnorm.exp fits a normal+exponential distribution to a specified data vector using maximum likelihood.

Description

Distrubtion function devined by: alpha*Normal(mean, varience)+(1-alpha) *Exponential(lambda).

Usage

1
2
Rnorm.exp(xi, wi = rep(1, NROW(xi)), guess = c(0.5, 0, 1, 1),
  tol = sqrt(.Machine$double.eps), maxit = 10000)

Arguments

xi

A vector of observations, assumed to be real numbers in the inveraval (-Inf,+Inf).

wi

A vector of weights. Default: vector of repeating 1; indicating all observations are weighted equally. (Are these normalized internally?! Or do they have to be [0,1]?)

guess

Initial guess for paremeters. Default: c(0.5, 0, 1, 1).

tol

Convergence tolerance. Default: sqrt(.Machine$double.eps).

maxit

Maximum number of iterations. Default: 10,000.

Details

Fits nicely with data types that look normal overall, but have a long tail starting for positive values.

Value

Returns a list of parameters for the best-fit normal distribution (alpha, mean, varience, and lambda).

Author(s)

Charles G. Danko


Questions? Problems? Suggestions? or email at ian@mutexlabs.com.

Please suggest features or report bugs with the GitHub issue tracker.

All documentation is copyright its authors; we didn't write any of that.