Package: EpiResTest
Type: Package
Title: Latent Residuals Test for Epidemiology and Ecology Models
Version: 1.0
Date: 2014-07-07
Author: Max S. Y. Lau; Jeffrey Pollock
Maintainer: Max S. Y. Lau <maxlauhk54@gmail.com>
Description: Latent residual tests recently developed (Lau et al. 2014) which
    are specifically designed to measure the goodness-of-fit of different model
    components of a general spatial S-E-I-R model commonly used in epidemiology
    and ecology studies. More restrictive model classes such as S-I-R can be
    accommodated. Note that although each test is specifically designed to be
    sensitive to one particular component of a spatio-temporal model, they are
    not theoretically restricted to only testing these specific aspects and can
    be used as a general model assessment tool like any other conventional
    model selection techniques. The underlying functions are coded in C++ so
    they should be generally quick.
Note: Under the null hypothesis that the model is correct, the
        residuals follow distribution U(0,1). The examples only compute
        a single set of residuals given actual values of most of the
        model components which are typically required to be imputed in
        data augmentation schemes such as MCMC in which multiple sets
        of residuals can be imputed. Only the raw residuals are
        outputted, a natural summary statistics of the residual samples
        is the p-value which may be computed using Kolmogorov-Smirnov
        or Anderson-Darling test.
License: GPL (>= 2)
Imports: Rcpp (>= 0.11.2)
LinkingTo: Rcpp, RcppArmadillo
Suggests: RUnit, ADGofTest
Packaged: 2014-12-05 14:12:57 UTC; sl451
NeedsCompilation: yes
Repository: CRAN
Date/Publication: 2014-12-05 15:29:39
