Vector generalized additive models.

Objects can be created by calls of the form `vgam(...)`

.

`nl.chisq`

:Object of class

`"numeric"`

. Nonlinear chi-squared values.`nl.df`

:Object of class

`"numeric"`

. Nonlinear chi-squared degrees of freedom values.`spar`

:Object of class

`"numeric"`

containing the (scaled) smoothing parameters.`s.xargument`

:Object of class

`"character"`

holding the variable name of any`s()`

terms.`var`

:Object of class

`"matrix"`

holding approximate pointwise standard error information.`Bspline`

:Object of class

`"list"`

holding the scaled (internal and boundary) knots, and the fitted B-spline coefficients. These are used for prediction.`extra`

:Object of class

`"list"`

; the`extra`

argument on entry to`vglm`

. This contains any extra information that might be needed by the family function.`family`

:Object of class

`"vglmff"`

. The family function.`iter`

:Object of class

`"numeric"`

. The number of IRLS iterations used.`predictors`

:Object of class

`"matrix"`

with \(M\) columns which holds the \(M\) linear predictors.`assign`

:Object of class

`"list"`

, from class`"vlm"`

. This named list gives information matching the columns and the (LM) model matrix terms.`call`

:Object of class

`"call"`

, from class`"vlm"`

. The matched call.`coefficients`

:Object of class

`"numeric"`

, from class`"vlm"`

. A named vector of coefficients.`constraints`

:Object of class

`"list"`

, from class`"vlm"`

. A named list of constraint matrices used in the fitting.`contrasts`

:Object of class

`"list"`

, from class`"vlm"`

. The contrasts used (if any).`control`

:Object of class

`"list"`

, from class`"vlm"`

. A list of parameters for controlling the fitting process. See`vglm.control`

for details.`criterion`

:Object of class

`"list"`

, from class`"vlm"`

. List of convergence criterion evaluated at the final IRLS iteration.`df.residual`

:Object of class

`"numeric"`

, from class`"vlm"`

. The residual degrees of freedom.`df.total`

:Object of class

`"numeric"`

, from class`"vlm"`

. The total degrees of freedom.`dispersion`

:Object of class

`"numeric"`

, from class`"vlm"`

. The scaling parameter.`effects`

:Object of class

`"numeric"`

, from class`"vlm"`

. The effects.`fitted.values`

:Object of class

`"matrix"`

, from class`"vlm"`

. The fitted values. This is usually the mean but may be quantiles, or the location parameter, e.g., in the Cauchy model.`misc`

:Object of class

`"list"`

, from class`"vlm"`

. A named list to hold miscellaneous parameters.`model`

:Object of class

`"data.frame"`

, from class`"vlm"`

. The model frame.`na.action`

:Object of class

`"list"`

, from class`"vlm"`

. A list holding information about missing values.`offset`

:Object of class

`"matrix"`

, from class`"vlm"`

. If non-zero, a \(M\)-column matrix of offsets.`post`

:Object of class

`"list"`

, from class`"vlm"`

where post-analysis results may be put.`preplot`

:Object of class

`"list"`

, from class`"vlm"`

used by`plotvgam`

; the plotting parameters may be put here.`prior.weights`

:Object of class

`"matrix"`

, from class`"vlm"`

holding the initially supplied weights.`qr`

:Object of class

`"list"`

, from class`"vlm"`

. QR decomposition at the final iteration.`R`

:Object of class

`"matrix"`

, from class`"vlm"`

. The**R**matrix in the QR decomposition used in the fitting.`rank`

:Object of class

`"integer"`

, from class`"vlm"`

. Numerical rank of the fitted model.`residuals`

:Object of class

`"matrix"`

, from class`"vlm"`

. The*working*residuals at the final IRLS iteration.`ResSS`

:Object of class

`"numeric"`

, from class`"vlm"`

. Residual sum of squares at the final IRLS iteration with the adjusted dependent vectors and weight matrices.`smart.prediction`

:Object of class

`"list"`

, from class`"vlm"`

. A list of data-dependent parameters (if any) that are used by smart prediction.`terms`

:Object of class

`"list"`

, from class`"vlm"`

. The`terms`

object used.`weights`

:Object of class

`"matrix"`

, from class`"vlm"`

. The weight matrices at the final IRLS iteration. This is in matrix-band form.`x`

:Object of class

`"matrix"`

, from class`"vlm"`

. The model matrix (LM, not VGLM).`xlevels`

:Object of class

`"list"`

, from class`"vlm"`

. The levels of the factors, if any, used in fitting.`y`

:Object of class

`"matrix"`

, from class`"vlm"`

. The response, in matrix form.`Xm2`

:Object of class

`"matrix"`

, from class`"vlm"`

. See`vglm-class`

).`Ym2`

:Object of class

`"matrix"`

, from class`"vlm"`

. See`vglm-class`

).`callXm2`

:Object of class

`"call"`

, from class`"vlm"`

. The matched call for argument`form2`

.

Class `"vglm"`

, directly.
Class `"vlm"`

, by class `"vglm"`

.

- cdf
`signature(object = "vglm")`

: cumulative distribution function. Useful for quantile regression and extreme value data models.- deplot
`signature(object = "vglm")`

: density plot. Useful for quantile regression models.- deviance
`signature(object = "vglm")`

: deviance of the model (where applicable).- plot
`signature(x = "vglm")`

: diagnostic plots.- predict
`signature(object = "vglm")`

: extract the additive predictors or predict the additive predictors at a new data frame.`signature(x = "vglm")`

: short summary of the object.- qtplot
`signature(object = "vglm")`

: quantile plot (only applicable to some models).- resid
`signature(object = "vglm")`

: residuals. There are various types of these.- residuals
`signature(object = "vglm")`

: residuals. Shorthand for`resid`

.- rlplot
`signature(object = "vglm")`

: return level plot. Useful for extreme value data models.- summary
`signature(object = "vglm")`

: a more detailed summary of the object.

Yee, T. W. and Wild, C. J. (1996)
Vector generalized additive models.
*Journal of the Royal Statistical Society, Series B, Methodological*,
**58**, 481--493.

```
# NOT RUN {
# Fit a nonparametric proportional odds model
pneumo <- transform(pneumo, let = log(exposure.time))
vgam(cbind(normal, mild, severe) ~ s(let),
cumulative(parallel = TRUE), data = pneumo)
# }
```

Run the code above in your browser using DataLab