FiniteDifferenceStaggeredStencilDerivation must be applied to field access

See merge request !109

Update AssignmentCollection.__repr__

See merge request !110

`Assignment Collection for y[0,0,0]` is usually not very helpful.

New representation:

```
In [6]:     forward_assignments = pystencils.AssignmentCollection({
   ...:         z[0, 0]: x[0, 0] * sympy.log(x[0, 0] * y[0, 0]),
   ...:     y[0,0] : x[1,1] +1
   ...:             })

In [7]: forward_assignments
Out[7]: AssignmentCollection: z_C, y_C <- f(x_C, x_NE)
```

otherwise the index gets lost

AssignmentCollection.__bool__

See merge request !108

Allow `CUSTOM_FIELD`s to have a different size in transformations.py

See merge request !107

This yields better error messages

AssignmentCollection can be used in many cases where you could also use
a List[Assignment]. With AssignmentCollection.__bool__, an empty
AssignmentCollection is falsy and a non-empty one truthy.

So you can `assert assignments, 'must not be emtpy'`

Fix FiniteDifferenceStaggeredStencilDerivation on Sympy 1.1

See merge request !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 !104

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

See merge request !102

Fix printing for Functions without name

See merge request !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 !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)