Fix FiniteDifferenceStaggeredStencilDerivation on Sympy 1.1

See merge request pycodegen/pystencils!105

Field.Access compare fails when one field access returns the offsets as ints in its hashable contents
and another one as sp.Integers.

Minor improvements to FiniteDifferenceStaggeredStencilDerivation

See merge request pycodegen/pystencils!104

Fix OpenCL for CUDA OpenCL by avoid int(...) constructor

See merge request pycodegen/pystencils!102

Fix printing for Functions without name

See merge request pycodegen/pystencils!103

This fixes test for LLVM GPU using pointer_arith_func

Prefer c-style cast instead.

finite difference stencil derivation for staggered positions

See merge request pycodegen/pystencils!99

Auto for assignments

See merge request pycodegen/pystencils!95

uses new Field.neighbor_vector

Reimplement create_staggered_kernel

Closes #16

See merge request pycodegen/pystencils!93

move_constants_before_loop needed to be changed to move already-existing variables up if necesesary and to preserve constness when renaming variables

Add __getnewargs__ to TypedImaginaryUnit

See merge request pycodegen/pystencils!94

Add own implementation for printing an unknown function

See merge request pycodegen/pystencils!96

also add a test for subexpressions

Using the SymPy implementation for printing a `sympy.Function` has some issues:

 - you cannot use `cast_func` as an argument -> SymPy prints
 `cast_func(2,float)
 - typed numbers are not typed any more -> e.g. pow(x, 0.25) instead of
 pow(x, 0.25f)

There was a problem when pickling TypedImaginaryUnit since it was using the __getnewargs__ of TypedSymbol.

More staggered grid improvements

See merge request pycodegen/pystencils!92