So I switched to eigenpy. But now I can't call functions from Python that require an Eigen:: container in the C++, because minieigen is removed and it had provided this layer.
But minieigen didn't work with numpy, a bigger failure. I think this barrier is smaller than adapting minieigen to numpy.
I started an issue over at Eigenpy, including listing some (at least partially) relevant SO posts.
So I switched to eigenpy. But now I can't call functions from Python that require an Eigen:: container in the C++, because minieigen is removed and it had provided this layer.
But minieigen didn't work with numpy, a bigger failure. I think this barrier is smaller than adapting minieigen to numpy.
I started an issue over at Eigenpy, including listing some (at least partially) relevant SO posts.