Tests facility

[szaghi]

After the pre-alpha development we need to design a suitable tests facilities. The library is aimed to be a general approximation (interpolation) tool thus the range of applications is quite wide. Nevertheless, one absolutely necessary test is its usage into the non linear PDEs solving framework. Due to my background, I will add a very simple 1D Riemann solver of compressible Euler PDEs system for testing the library: this should be simple enough to be a yet clear test while it retains a lot complexity (unsteady non linear hyperbolic system).

Here I would like to discuss and list all possible tests.

Tests list

• [ ] 1D Riemann solver for the compressible Euler system;
  • [ ] Shu-Osher problem;
• [ ] image interpolation (procedure for increasing resolution)?
• [ ] check conservative feature of the interpolations when coupled with (fully) conservative schemes;
• [x] reconstruct the sin trigonometric fuction;

[giacrossi]

Another important test, in my opinion, is to verify that the interpolation is fully conservative, in addition to be very accurate and non-oscillatory. Honestly, I don't know what pratical test can be used to obtain this goal.

[szaghi]

@giacombum The Gay-Lussac Riemann problem is commonly used for checking conservation errors long running unsteady simulations. Added to list! Thank you very much!

[zbeekman]

Shu-Osher problem for testing oscilations:

[szaghi]

@zbeekman sure, there are a lot of 1D RP tests we must implement as "regression-tests-suite". Shu-Osher one added to the list! Thank you zaak