Description Usage Arguments Details Value See Also Examples
View source: R/nauf_trms_fr_x_z.R
nauf_model.frame
creates a model frame which employs
nauf_contrasts
for unordered factors.
1 2 3 |
formula, data, subset, ... |
See |
na.action, drop.unused.levels, xlev, contrasts |
Changes from default values for these arguments are ignored with a warning. |
ncs_scale |
A positive number passed as the |
First, the default method for model.frame
is called.
Then any variable in the resulting model frame that is not an unordered
factor but has only two unique non-NA
values is coerced to an
unordered factor. Unordered factors are then assigned contrasts with
named_contr_sum
, passing ncs_scale
as the
function's scale
argument. Then, necessary contrast changes in
interaction terms and random effects slopes are determined as described in
nauf_contrasts
.
The recommended usage is to first standardize
the
regression variables, and then use the formula
and data
elements in the resulting standardized
object as arguments to
nauf_model.frame
. When this is done, ncs_scale
is obtained
from the standardized.scale
attribute of the formula
, unless
ncs_scale
is specified as a value which does not match the
standardized
scale, in which case the explicitly specified
ncs_scale
argument is used with a warning. If
standardize
is not used prior to calling
nauf_model.frame
, then ncs_scale
defaults to 1
unless
explicitly specified in the function call, in which case the specified value
is used.
Changes from the following default values are ignored with a warning:
This default value is required in order for
NA
values to be treated as defined in
nauf_contrasts
.
This default value is set because
nauf_model.frame
assumes that data
is not new data. To
create a nauf.frame
with new data, the terms
attribute of an already existing nauf.frame
(which
has class nauf.terms
) can be used as the
formula
argument to model.frame
.
This default is necessary for the same reasons as the
default value for drop.unused.levels
.
For unordered factors, contrasts are automatically
created with named_contr_sum
, as sum contrasts
are necessary to implement nauf_contrasts
. To specify
custom contrasts for ordered factors, the custom contrasts should be
explicitly assigned to the ordered factor in data
(this is
automatically done if standardize
is used
first as recommended).
A model frame with second class attribute nauf.frame
. Its
formula
attribute has class nauf.formula
and its terms
attribute has class nauf.terms
.
nauf_contrasts
for a description of the contrasts
applied to unordered factors, nauf_model.matrix
for obtaining
a fixed effects model matrix, and nauf_glFormula
for
obtaining both fixed effects and random effects model matrices.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 | dat <- plosives
dat$spont[dat$dialect == "Valladolid"] <- NA
form <- intdiff ~ voicing * dialect * spont +
(1 + voicing * spont | speaker) + (1 + dialect | item)
## default behavior when standardize is not used
# defaults to ncs_scale = 1
mf <- nauf_model.frame(form, dat)
# uses specified ncs_scale = 0.5
mf_0.5 <- nauf_model.frame(form, dat, ncs_scale = 0.5)
## standardize first (recommended use)
sobj <- standardize(form, dat)
sobj_0.5 <- standardize(form, dat, scale = 0.5)
# uses ncs_scale = 1 from attr(sobj$formula, "standardized.scale")
mf_sobj <- nauf_model.frame(sobj$formula, sobj$data)
# uses ncs_scale = 0.5 from attr(sobj_0.5$formula, "standardized.scale")
mf_sobj_0.5 <- nauf_model.frame(sobj_0.5$formula, sobj_0.5$data)
## Not run:
## not recommended
# uses specified ncs_scale = 0.5 and issues a warning since
# attr(sobj$formula, "standardized.scale") = 1
mf_warning <- nauf_model.frame(sobj$formula, sobj$data, ncs_scale = 0.5)
## End(Not run)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.