- small (length < 5) arrays with shape and stride information had to be
  memcpy'd to the GPU before every kernel call
- instead of passing the information as arrays, the single elements are
  passed
- leads to more function arguments, but simplifies GPU kernel calls

-> changes in all backends required

- correction functions

- currently scalar multiplication (field-access * Matrix) does not work
  since sympy classifies field-accesses as iterable and tries matrix
  multiplication instead of scalar multiplication

- Added test for linear equation system and moved it to pystencils.utils

- new functions to easily extract stencil coefficients and visualize them
- Moved stencil functions from lbmpy to pystencils

- simpler function to init N phase model

- 1D benchmark scenario, trying different correction functions
- 2D neumann angle evaluation based on circle fitting

- simplified pressure tensor formulas
- added test that checks if force is mathematically equal if computed from pressure tensor or directly

-> test_numerical_1D notebook

- CSE has to be done after dynamic relaxation rate adaption

- separate lbmpy module
- code generation for walberla
- isolated code to generate a listlbm update rule from collision rule

- these modifications have been made previously in the update rule
- update rule should only be about field access, not method modifications

- defined ordering of headers
- lbm split kernels produced different ordering of assignments due to
  usage of a set

- fix for field collection and reduction operations
- support for aligned fields in parallel data handling

- new time loop caches all kernel functions with their argument dict
  -> inner loop just calls C functions with cached kwargs