singleRcapture 0.1.0
Pre-releaseThe first version of the singleRcapture package that is somewhat suitable for use.
Features:
-
Zero truncated count data regression with Poisson, Geometric and Negative Binomial models and their application to population size estimate, analytic, nonparametric and parametric bootstrap variance estimations are available. Two methods of constructing analytic confidence intervals are supported studentised and by logarithmic transformation.
-
Zero truncated count data regression with Poisson, Geometric and Negative Binomial models and their application to population size estimate, analytic and non parametric bootstrap variance estimations are available.
-
Two mixture models, chao and zelterman, utilising logit regression for population size estimation, analytic and nonparametric variance estimations are available.
-
The above models are applied with their respective
family
class functions (model
parameter) that are passed to main functionestimate_popsize
. -
Two fitting algorithms,
IRLS
and simple MLE byoptim
. -
summary
method for the model. It is handled differently insingleRcapture
than in glm's. -
marginalFreq
function for creating a table of marginal frequencies. -
summary
method for marginal frequencies that applies chisq and G tests for goodness of fit -
vcov
function for extracting the covariance matrix of regression parameters. -
basic
control.pop.var
parameters for variance estimation.