meanexp.conditional: Locating the parameter of an exponentially distributed 'X'...

In arnejohannesholmin/AJ: Holmins functions, used also for loading all packages by library(AJ)

Description

Locating the parameter of an exponentially distributed 'X' based on the mean of Y=X|X<C or Y=X|X>C. Limited accuracy if the mean is close to the limiting value of C/2 (for which the distribution of the lower values is almost uniform). 2012-05-23: The function returns biased estimates for the case type=="l", especially for small sample sizes. The reason for this is not yet established. Use the function meanexp.quantile() instead.

Usage

meanexp.conditional(mean, C, x = NULL, beta = C, pres = C * 1e-04,
  max.iter = 100, type = "l")

Arguments

mean	is the mean of the values of the exponentially distributed lower or greater than 'C'.
C	it the value defining 'Y'.
x	it a vector of the data from which 'mean' is calculated (overrides 'mean' if given).
beta	is the initial parameter value, defaulted to 'C'. Should not be chosen larger than 'C'.
pres	is the required presicion of the iteration.
max.iter	is the maximum number of iterations.
type	is "l" if 'mean' is the mean of X|X<C and "g" if 'mean' is the mean of X|X>C.

arnejohannesholmin/AJ documentation built on May 27, 2019, 4:06 a.m.