Dr. Bangerth,
Thanks so much for your reply that pointed out my problem.
Quadrature is applied to the residual expression, which includes both
analytical and approximation.
"right-hand-side" showed me that approximation can be obtained at
quadrature point, before getting Jacobians and weighted.
Good morning! Dr. Bangerth,
Thanks so much for your reply, this is wonderful!
Only approximation needs to do quadrature, analytical soln. only has
quadrature points.
I will keep testing today.
Best,
Judy
On Wednesday, December 22, 2021 at 12:27:59 AM UTC-5 Wolfgang Bangerth
wrote:
> On 12/21/21
On 12/21/21 8:59 AM, Judy Lee wrote:
Can I make an expression u(x) for 1D analytical solution, that can take
*q_index* for evaluating u(@ *q_index*), while skip obtaining *q_index*
locations/points value?
What is 'q_index'? It's usually the index (an integer) of a quadrature point
on a cell
Hello everyone!
Can I make an expression u(x) for 1D analytical solution, that can take
*q_index* for evaluating u(@ *q_index*), while skip obtaining *q_index*
locations/points value?
I am doing the residual: *u(x) - u_h(x)* under each cell, in which *u_h(x)*
comes from approximation that has
