
Gets various estimates from an ic_np
, ic_sp
or ic_par
object.
# S3 method for icenReg_fit
predict(object, type = "response", newdata = NULL, ...)
Model fit with ic_par
or ic_sp
type of prediction. Options include "lp", "response"
data.frame
containing covariates
other arguments (will be ignored)
Clifford Anderson-Bergman
If newdata
is left blank, will provide estimates for original data set.
For the argument type
, there are two options. "lp"
provides the
linear predictor for each subject (i.e. in a proportional hazards model,
this is the log-hazards ratio, in proportional odds, the log proporitonal odds),
"response"
provides the median response value for each subject,
*conditional on that subject's covariates, but ignoring their actual response interval*.
Use imputeCens
to impute the censored values.
simdata <- simIC_weib(n = 500, b1 = .3, b2 = -.3,
inspections = 6,
inspectLength = 1)
fit <- ic_par(cbind(l, u) ~ x1 + x2,
data = simdata)
imputedValues <- predict(fit)
Run the code above in your browser using DataLab