problem with interpolated function in eoc loop
Expected behaviour: Upon running the code, the "H-1 error" should converge with an eoc of about 2. This is observable by running the loops separately and calculating the errors manually.
What happens: On subsequent loops, the error diverges by a large amount.
The error seems to be fixable by adding:
uh.interpolate(exact)
when inside the loop, but retaining
uh = space2.interpolate(exact, name="uh")
on the outside. However this is a bug since having the second line in the loop shouldn't cause it to fail.
Edited by Lloyd Connellan