Reproduces the "Resids vs. linear pred" plot from
gam.check but using randomised quantile residuals, a la Dunn and Smyth (1996). Checks for heteroskedasticity as as usual, looking for "funnel"-type structures in the points, which is much easier with randomised quantile residuals than with deviance residuals, when your model uses a count distribution as the response.
arguments passed on to all plotting functions
Note that this function only works with negative binomial and Tweedie response distributions.
Earlier versions of this function produced the full
gam.check output, but this was confusing as only one of the plots was really usedul. Checks of
k are not computed, these need to be done using
Based on code provided by Natalie Kelly, bugs added by Dave Miller
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
library(Distance) library(dsm) library(tweedie) # load the Gulf of Mexico dolphin data (see ?mexdolphins) data(mexdolphins) # fit a detection function and look at the summary hr.model <- ds(distdata, max(distdata$distance), key = "hr", adjustment = NULL) # fit a simple smooth of x and y with a Tweedie response with estimated # p parameter mod1 <- dsm(count~s(x, y), hr.model, segdata, obsdata, family=tw()) rqgam.check(mod1)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.