Roman

Add Vc::pow and Vc::cbrt functions

3

Some math functions like logarithm and trigonometric ones already exist for `float_v` and `double_v`. Would be a good idea to add a portable `Vc::pow` function as well. Is there any...

Implement a fast SVD routine

3

At the moment Fastor provides the fastest LU decomposition, inversion and solve routines for small tensors but lacks a fast SVD routine

Investigate overloading operator() with std::initializer_list

2

The operator() for tensors could be overloaded to provide a similar functionality to `seq`, `fseq` and `iseq`, like so ~~~c++ operator()(std::initializer_list _s1, std::initializer_list _s2); ~~~ The tensor could then be...

Implement a fast eigen decomposition

1

At the moment Fastor provides the fastest LU decomposition, inversion and solve routines for small tensors but lacks a fast eigen decomposition routine

All tensor types and their `assign/trivial_assign` functions should use `is_aligned` instead of `FASTOR_ALIGNED`. 1. We should introduce `alignment_t` or `alignment_v` for all tensor types with storage 2. We should introduce...

Specialisation for SIMDVector<bool,ABI>?

2

Just opening for discussion. But implementing masked vectors for each vector type *might* give some performance advantage over the current implementation

Importrant is PEP8. PEP257 is mainly already in place, although some functions lack docstrings.