Is it because in much of the maths domain it turns out you manipulate columns more often than rows?
(not a mathematician but I could believe if the columns represent "dimensions" or "qualities" of some dataset, and you want to apply a function to a given dimension, having data in column-natural order makes that faster.)
Obviously you want to believe naievely there is no difference to the X and Y in the X,Y plane, but machines are not naieve and sometimes there IS a difference to doing things to the set of verticals, and the set of horizontals.
Is it because in much of the maths domain it turns out you manipulate columns more often than rows?
(not a mathematician but I could believe if the columns represent "dimensions" or "qualities" of some dataset, and you want to apply a function to a given dimension, having data in column-natural order makes that faster.)
Obviously you want to believe naievely there is no difference to the X and Y in the X,Y plane, but machines are not naieve and sometimes there IS a difference to doing things to the set of verticals, and the set of horizontals.