getRank: Ranking based on Estimates

In ordBTL: Modelling comparison data with ordinal response

Description

Extracts the estimated parameters and sorts them based on their estimated values

Usage

getRank(ordBTL, decreasing=TRUE, prefix=NULL, reference=NULL)

Arguments

ordBTL	a fitted model returned by ordBTL.
decreasing	logical. Should the sort be increasing or decreasing?
prefix	(optional) a character that is included in the names of the parameters; only the parameters are returned that include this character (prefix=NULL extracts all estimated parameters).
reference	(optional) a character specifying the reference object.

Value

matrix containing the parameter estimates.

Author(s)

Giuseppe Casalicchio

See Also

ordBTL

Examples

# Get the CEMS data and generate design matrix
example(wide2long, package="ordBTL", echo=FALSE)
des1 <- design(CEMSlong, var1="object1", var2="object2", 
               use.vars="Y", reference="Stockholm")

# Fit the adjacent categories model, which corresponds to 
# the log-linear BTL model (see Agresti, 1992)
mod1 <- ordBTL(Y~., data=des1, family="acat", 
               family.control=list(reverse=TRUE))

# Extract all parameter estimates
getRank(mod1)

# Extract all parameter estimates and add parameter for
# reference object (which is set to zero)
getRank(mod1, reference="GAMMA.Stockholm")

# Extract only parameter estimates that include the 
# string "Intercept"
getRank(mod1, prefix="Intercept")

# Extract only parameter estimates that include the 
# string "GAMMA" (which will be the object parameters)
getRank(mod1, prefix="GAMMA")

ordBTL documentation built on May 2, 2019, 7:59 a.m.