Calculate various summaries of smoothing spline ANOVA fits with non-Gaussian responses.
# S3 method for gssanova0
summary(object, diagnostics=FALSE, ...)summary.gssanova0 returns a list object of class
"summary.gssanova0" consisting of the following elements.
The entries pi, kappa, cosines, and
roughness are only calculated if diagnostics=TRUE.
Fitting call.
Error distribution.
Method for smoothing parameter selection.
Assumed or estimated dispersion parameter.
Number of performance-oriented iterations performed.
Fitted values on the link scale.
Working residuals on the link scale.
Residual sum of squares.
Deviance residuals.
Deviance of the fit.
Deviance of the null model.
Estimated size for family="nbinomial" with one
column responses. Estimated inverse scale of log life time for
family="nbinomial", "lognorm", or
"loglogis".
Roughness penalty associated with the fit.
"Percentage decomposition" of "explained variance" into model terms.
Concurvity diagnostics for model terms. Virtually the square roots of variance inflation factors of a retrospective linear model.
Cosine diagnostics for practical significance of model terms.
Percentage decomposition of the roughness penalty
penalty into model terms.
Object of class "gssanova".
Flag indicating if diagnostics are required.
Ignored.
Similar to the iterated weighted least squares fitting of
glm, penalized likelihood regression fit can be calculated
through iterated penalized weighted least squares.
The diagnostics are based on the "pseudo" Gaussian response model behind the weighted least squares problem at convergence.
Gu, C. (1992), Diagnostics for nonparametric regression models with additive terms. Journal of the American Statistical Association, 87, 1051--1058.
Fitting function gssanova0 and methods
predict.ssanova0, fitted.gssanova.