It generates a matrix of basis functions using local linear splines. The number of basis functions and the breaking points are determined by the knot specifications of the slseModel or cslseModel model.
## S3 method for class 'slseModel'llSplines(object,...)## S3 method for class 'cslseModel'llSplines(object,...)
Arguments
object: A model of class slseModel or cslseModel.
...: Additional arguments to pass to other methods. Currently not used.
Returns
The function returns a matrix of basis functions used to estimate the semiparametric model. It is used directly as regressor in lm as shown in the example below.
Examples
data(simDat3)## We manually estimate the semiparametric modelmod1 <- slseModel(Y ~ X1 * X2, data = simDat3)U <- llSplines(mod1)fit1 <- lm(Y ~ U, data = simDat3)## We use estSLSE instead (results are identical)fit2 <- estSLSE(mod1)