estimate.rds3: Main function in rds2 package. Returns the ML estimate of...

In rds2: Population size and degree distribution estimation in respondent driven sampling.

Description

Performs maximum likelihood estimation of the population size, degree distribution and theta [explain] allowing (a) "soft" constraints on theta and (b) an input Sij matrix.

Usage

  estimate.rds3(data, Sij, init, const, arc = FALSE,
    maxit = 10000, initial.thetas, theta.minimum,
    theta.range)

Arguments

data	An integer vector of the degrees of each individual sampled.
Sij	An integer matrix of the counts of individuals with rank j at sampling period t. Row names correspond to the degrees.
init	Optional initialization values.
const	Control the concentration of the mapping of the theta parameter from the contstrained domain to the real line.
arc	Should the degrees in the snowball be taken in account (TRUE) or just the size of the snowball (FALSE). Defaults to FALSE.
maxit	Number of maximal optimization iterations.
initial.thetas	Vector of initial values of theta parameter.
theta.minimum	Lower bound on the allowed values of theta.
theta.range	Allowed range of theta values. From theta.minimum to theta.minimum+range.

Author(s)

Jonathan Rosenblatt

Examples

data(simulation)
temp.data<- unlist(data3[1,7000:7500])
estimate.rds3(temp.data, Sij = make.Sij(temp.data), initial.thetas = c(1,10), arc = FALSE, maxit = 1000, const = 0.5, theta.minimum = -0.5, theta.range = 2)

rds2 documentation built on May 2, 2019, 6:42 p.m.