The function removes dimensions from a tensor by summing all entries which only differ in these dimensions.
margin.tensor(X,i=NULL,by=NULL)
the tensor
the dimensions to be removed
instead of i the dimensions to be kept
The tensor with all elements only differing only in the dimensions specified added up and only the other dimensions left over.
This is a tensor multiplication with the \(1_i\) tensor.
to.tensor
# NOT RUN { A <- diag(1:5) A margin.tensor(A,1) A <- to.tensor(1:30,dim=c(i=3,j=5,k=2)) ftable(A) margin.tensor(A,"j") # }
Run the code above in your browser using DataLab