inspect.glm {asuR} | R Documentation |
This functions can be used to validate some assumptions of generalized linear models.
## S3 method for class 'glm': inspect(mymodel, which = c("select", "sequence", "all"), id = c("all", "none"), ...)
mymodel |
an object of class glm , usually the result of a
call to the function glm |
which |
a character string; do you want to select the plots from
a list of all possible diagnostics plots available or do you want
all of them, one after the other |
id |
a character string or numeric value; in which panel should it be possible to interactively identify values |
... |
further arguments |
For glm available:
description | short & | long function names |
deviance residuals vs. linear predictor | dep | devianceResidual.linearPredictor |
partial residual vs. each predictor | rpp | partialResidual.eachPredictor |
linearized response vs. linear predictor | lep | linearizedResponse.linearPredictor |
half-normal quantiles vs. absolute stud. residuals | hnp | halfNormalQuantiles.absoluteStudentizedResiduals |
A list with identified values (row indices), one named slot (short name) for each inspection function and an additional slot (called all) with all values that were selected at lest in one plot.
This functions help to inspect glm's with continuous predictors, sometimes they are anyway informative also for categorical predictors ...
plot
## data(gala) ## model <- glm(Species ~ log(Area) + log(Elevation) + log(Nearest) + log(Scruz+0.1) + log(Adjacent), family=poisson(link=log), data=gala) ## inspect(model)