I am working on some B-spline regression (first degree, single knot, not very complicated) and would like to compare the parameter estimates for the portion before and the portion after the internal knot.
Right now I have something like this:
fit <- lm(y ~ bs(x, degree = 1, knots = 20), data = bar)
The fit then has the intercept estimate and two sets of parameter estimates.
I am interested in comparing those two parameter sets against one an other - does anyone know to divide the bs model object or extract those two sub-models? Or, convert an F value to a p value in R so I can do those tests manually.
