Comment #24 on issue 2225 by mrock...@gmail.com: Enhancing Matrix Norm
http://code.google.com/p/sympy/issues/detail?id=2225
Regarding Frobenius vs Spectral/2-norm
Mathematically the spectral/2-norm is "the" standard norm just like the
euclidian/2-norm is standard for vectors. Vector norms have matching Matrix
norms. Spectral matches Euclidian.
However - Both the current computation of eigenvalues and the Max issue
makes computing the spectral norm very challenging and fail for many cases.
Maybe this is a case when it's better to return the Frobenius norm which is
no more complex than the terms in the original matrix and always computes.
The people who care about Matrix norms will know to check. Most of us know
that you need to check after having used Numpy (which has Frobenius as
default). Or maybe we don't have a default?
Having a lazy Norm operator is interesting. Is there consensus in SymPy as
to which operators should be lazy (evaluated on request) and which are
calculated immediately?
--
You received this message because you are subscribed to the Google Groups
"sympy-patches" group.
To post to this group, send email to sympy-patches@googlegroups.com.
To unsubscribe from this group, send email to
sympy-patches+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/sympy-patches?hl=en.