FEMCE

Now using sparse matrices, it takes 1000 times less RAM to run. What took 7 GB now takes 7 MB. The solver is also a lot faster thanks to this, and now the parallelized version of the code also scales linearly with the number of mesh elements. Users can now run simulations with millions of elements.

Minor updates on the graphical representation of the magnetic field. Scatter plots are more efficient for very refined meshes.

Minor update fixing the rotating magnetocaloric effect which was crashing the app.

Implemented the changes required to implement C into FEMCE through mex, on its next release.