Convert Between Formula and Index
These functions will convert a formula object to a binary index and vice versa.
formula2index(big.X, formula, data) index2formula(index, maximal.mod)
big.X |
The design matrix under the maximal model. |
formula |
An object of class |
data |
An object of class |
maximal.mod |
An object of class |
index |
A binary vector, of the same length as the number of log-linear parameters in the maximal model, indicating which parameters are present in the model to be converted to a formula. |
The function formula2index will produce a binary vector of the same length as the number of log-linear
parameters in the maximal model, indicating which parameters are present in the model represented by the
argument formula.
The function index2formula will produce an object of class "formula": a symbolic description of the model given by
the argument index.
This function will not typically be called by the user.
Antony M. Overstall A.M.Overstall@soton.ac.uk.
data(ScotPWID) ## Load the ScotPWID data maximal.mod<-glm(y~(S1+S2+S3+S4+Region+Gender+Age)^2,family=poisson,contrasts=list( S1="contr.sum",S2="contr.sum",S3="contr.sum",S4="contr.sum", Region="contr.sum",Gender="contr.sum",Age="contr.sum"),data=ScotPWID,x=TRUE) ## Fit the maximal model containing all two-way interactions. big.X<-maximal.mod$x ## Set the design matrix under the maximal model index<-formula2index(big.X=big.X, formula=~S1+S2+S3+S4+Region+Gender+Age+S1:S2+S1:Age+S2:Gender+S3:S4+S4:Age, data=ScotPWID) ## Find the index under the model with the following interactions: ## S1:S2 ## S1:Age ## S2:Gender ## S3:S4 ## S4:Age index ## Print index # [1] 1 1 1 1 1 1 1 1 1 0 0 0 0 1 0 0 0 1 0 1 0 0 0 0 0 1 0 0 0 index2formula(index=index,maximal.mod=maximal.mod) ## Go back to formula #y ~ S1 + S2 + S3 + S4 + Region + Gender + Age + S1:S2 + S1:Age + # S2:Gender + S3:S4 + S4:Age
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.