powered by
Using results calculated by the linearHypothesis function in the car package, calculate a linear combination of regression parameters.
linearHypothesis
car
lincom(model, specification, level = 0.95, eform = FALSE, ...)
A matrix with columns including the estimate, a normal-based confidence interval, test statistic and p-values.
regression model object (as per the model argument in linearHypothesis)
model
specification of the linear combination. This is the same as a single component of the hypothesis.matrix argument in linearHypothesis.
hypothesis.matrix
the confidence level required
logical for whether to exponentiate the confidence interval (default=FALSE)
other arguments to the linearHypothesis function.
Multiple specifications of linear combinations are called individually.
See Also linearHypothesis.
fit <- glm(chd ~ hieng*job + offset(log(y)), data=diet, family=poisson) lincom(fit, c("hienghigh+hienghigh:jobconductor", "hienghigh+hienghigh:jobbank"), eform=TRUE)
Run the code above in your browser using DataLab