predict.mvord: Predict method for Multivariate Ordinal Regression Models.

In mvord: Multivariate Ordinal Regression Models

Predict method for Multivariate Ordinal Regression Models.

Description

Obtains predicted or fitted values for objects of class 'mvord'.

Usage

## S3 method for class 'mvord'
predict(
  object,
  newdata = NULL,
  type = NULL,
  subjectID = NULL,
  newoffset = NULL,
  ...
)

Arguments

object	an object of class 'mvord'.
newdata	(optional) data frame of new covariates and new responses.
type	types "class", "prob" and "cum.prob" are available.
subjectID	(optional) vector specifying for which subjectIDs the predictions or fitted values should be computed.
newoffset	(optional) list of length equal to the number of outcomes, each element containing a vector of offsets to be considered.
...	further arguments passed to or from other methods.

Details

type	description
"class"	combination of response categories with the highest probability. Used as default if newdata contains no column(s) for response variable(s) or all responses are NA.
"prob"	fitted joint probability for the observed response categories
	or the categories provided in the response column(s) in newdata. Used as default if newdata contains column(s) for response variable(s).
	If response column(s) in newdata contain only NAs or is missing, this type is not supported.
"cum.prob"	fitted joint cumulative probability for the observed response
	categories or the categories provided in the response column(s) in newdata.
	If response column(s) in newdata contain only NAs or is missing, this type is not supported.

If provided, the (row) names of the output correspond to the subjectIDs, otherwise they correspond to the row id of the observations.

See Also

marginal_predict, joint_probabilities

mvord documentation built on Sept. 11, 2024, 7:21 p.m.