Hi,

I'm translating some MATLAB code to Eigen and I have something I would write in Eigen as:

T = B1.array().colwise() * c1.array() + B2.array().colwise() * c2.array() + ....

with Bs as matrixes and cs as vectors.

I seem to be able to compile

T = B1.array().colwise() + c1.array()

but not

T = B1.array().colwise() * c1.array()

is this pending to be implemented or I'm missing something?

Thanks
Martin

Indeed, AFAICS this isn't implemenented and would be a good thing to have; however you can already do the same with a diagonal product (which will be optimized just as well) assuming all your objects are matrices:

Code: Select all

T = c1.asDiagonal() * B1;


hi bjacob,

Thanks for your answer. it's a shame it's not implemented yet. I look forward to it

I'll use the expression you suggested.

Martin.