Detach the EHVec class from Matrix

[willGraham01]

Description

The EHVec class is currently a public Matrix<fftw_complex>. It is desirable to convert Matrix<> to used strided vector storage and remove any MATLAB dependencies.

The former doesn't work for EHVec since the fftw_complex type is a double[2], which throws an error when used in templated std::vector<T> functions. Additionally, EHVec is used in such a way that fftw needs a continuous buffer to perform the forward/backward transform, which a strided vector does not guarantee.

As such, it is necessary to provide EHVec with it's own (MATLAB-free) class. This enables Matrix and its other derivatives to be freed of MATLAB influence.

• Concerns Classes to detach from MATLAB #285, resolves the EHVec task.
• Depends on DispersiveMultiLayer no longer uses MATLAB #324, targets the expected state of main.

Testing

• Existing unit tests for EHVec continue to pass after the refactor.
• Removed the EHVecTest class since it is no longer a member of the arrays.h file, and it only had a single unit test in the first place (so a class was overkill).

[codecov]

Codecov Report

Patch coverage: 70% and no project coverage change.

Comparison is base (fdbfc93) 26% compared to head (4154d3d) 26%.

Additional details and impacted files

@@         Coverage Diff         @@
##           main   #303   +/-   ##
===================================
  Coverage    26%    26%           
===================================
  Files        69     71    +2     
  Lines      3674   3688   +14     
===================================
+ Hits        952    957    +5     
- Misses     2722   2731    +9     

Impacted Files	Coverage Δ
tdms/include/arrays.h	29% <ø> (-3%)	⬇️
tdms/include/arrays/eh_vec.h	0% <0%> (ø)
tdms/include/field.h	0% <ø> (ø)
...ms/include/simulation_manager/simulation_manager.h	0% <ø> (ø)
tdms/src/arrays.cpp	91% <ø> (-<1%)	⬇️
tdms/src/arrays/eh_vec.cpp	88% <88%> (ø)

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[samcunliffe]

Totally fine with this... but I'm a bit sad that I can't have a Matrix<std::complex>.

What was the error?

[willGraham01]

Totally fine with this... but I'm a bit sad that I can't have a Matrix<std::complex>.

What was the error?

Matrix<std::complex> doesn't complain - it's Matrix<fftw_complex> which does. This is because fftw_complex is a typedef of double[2], and then some obscure error that I think boils down to the fact that an fftw_complex is a double* so can't be used in the template.