Here, it's outer and inner. Boost ublas supports col- and row-major matrices in almost the same way as does Eigen and that's propably why they opted for post fixes '1' and '2'.
Ok, let me try to give you a full list of what boost ublas offers:
dense typedefs
matrix::iterator1 (model of 'Indexed Random Access Column/Row Iterator')
matrix::iterator2 (model of 'Indexed Random Access Column/Row Iterator')
vector::iterator (model of 'Indexed Random Access Iterator')
sparse typedefs:
sparse_matrix::iterator1 (model of 'Bidirectional Column/Row Iterator')
sparse_matrix::iterator2 (model of 'Bidirectional Column/Row Iterator')
sparse_vector::iterator (model of 'Indexed Bidirectional Iterator')
Then there exist the corresponding const_ typedefs.
The functions are (for dense and sparse matrices, const and non-const versions)
begin1()
end1()
begin2()
end2()
and (for vectors, also const and non-const)
begin()
end()
For the sake of completeness, they also provide for all iterator types the 'reverse' versions all prefixed with 'r'. As I said, it's a lot.
Here is the reference:
http://www.boost.org/doc/libs/1_41_0/libs/numeric/ublas/doc/index.htmRegards,
Hauke