This function is used to extract a coefficient from a fitted `pfr` model, in
particular smooth functions resulting from including functional terms specified
with lf, af, etc. It can also be used to extract smooths
genereated using mgcv's s, te, or t2.
# S3 method for pfr
coefficients(
object,
select = 1,
coords = NULL,
n = NULL,
se = ifelse(length(object$smooth) & select, TRUE, FALSE),
seWithMean = FALSE,
useVc = TRUE,
Qtransform = FALSE,
...
)# S3 method for pfr
coef(
object,
select = 1,
coords = NULL,
n = NULL,
se = ifelse(length(object$smooth) & select, TRUE, FALSE),
seWithMean = FALSE,
useVc = TRUE,
Qtransform = FALSE,
...
)
a data frame containing the evaluation points,
coefficient function values and optionally the SE's for the term indicated
by select.
return object from pfr
integer indicating the index of the desired smooth term
in object$smooth. Enter 0 to request the raw coefficients
(i.e., object$coefficients) and standard errors (if se==TRUE).
named list indicating the desired coordinates where the
coefficient function is to be evaluated. Names must match the argument names
in object$smooth[[select]]$term. If NULL, uses n
to generate equally-spaced coordinates.
integer vector indicating the number of equally spaced coordinates
for each argument. If length 1, the same number is used for each argument.
Otherwise, the length must match object$smooth[[select]]$dim.
if TRUE, returns pointwise standard error estimates. Defaults
to FALSE if raw coefficients are being returned; otherwise TRUE.
if TRUE the standard errors include uncertainty about
the overall mean; if FALSE, they relate purely to the centered
smooth itself. Marra and Wood (2012) suggests that TRUE results in
better coverage performance for GAMs.
if TRUE, standard errors are calculated using a covariance
matrix that has been corrected for smoothing parameter uncertainty. This
matrix will only be available under ML or REML smoothing.
For additive functional terms, TRUE indicates the
coefficient should be extracted on the quantile-transformed scale, whereas
FALSE indicates the scale of the original data. Note this is
different from the Qtransform arguemnt of af, which specifies
the scale on which the term is fit.
these arguments are ignored
Jonathan Gellar and Fabian Scheipl
Marra, G and S.N. Wood (2012) Coverage Properties of Confidence Intervals for Generalized Additive Model Components. Scandinavian Journal of Statistics.