I am converting some algorithms from Python/Numpy to Julia and have no idea 
how to handle that Julia arrays are column major.

I have a matrix on disk (very large > 10GB) and it is stored row major. 
This is the natural representation in my application.
Now I want to solve a linear system of equations and use the Kaczmarz 
algorithm that operates on matrix rows and therefore should be used with a 
row major matrix representation (Please no comments on the convergence 
speed of Kaczmarz algorithm, its extremely fast in my application).

I could write the matrix loading and the algorithm keeping the 
transposition in mind but this does not feel right to me.
Another solution seems to be the introduction of a cheap transposition 
type. But even more general one could make the Julia base array more 
flexible and allow for arbitrary strides.

Has someone run into similar issues?

Reply via email to