I think an example of how this BC handling works for spectral operators would be a good thing to add to the writeup. Essentially, you can write out Dirichlet and Neumann conditions as summation conditions in the function space (just plug in x and evaluate). The summation condition is just a row/column of the operator. This defines how you would have to extrapolate to a larger function space that satisfies the conditions, and then the differentiation operator brings it back down.
I think an example of how this BC handling works for spectral operators would be a good thing to add to the writeup. Essentially, you can write out Dirichlet and Neumann conditions as summation conditions in the function space (just plug in
x
and evaluate). The summation condition is just a row/column of the operator. This defines how you would have to extrapolate to a larger function space that satisfies the conditions, and then the differentiation operator brings it back down.