search

PythonOptimizers / cysparse

Python/Cython library to replace PySparse
http://PythonOptimizers.github.io/cysparse
7 stars 3 forks source link

What about conjugate and transposed conjugate proxies for ALL sparse matrices? #211

Closed ghost closed 8 years ago

ghost commented 8 years ago

For the moment, only complex matrices do have conjugate and a transposed conjugate proxies.

Do we still keep it that way? Or do real matrices return the identity for the conjugate and the transposed for the transposed conjugate?

Something like:

A.conj == A  # A is a real matrix
A.H == A.T   # A is a real matrix

Same question for matvec_htransp and matvec_transp.

What do you think?

@dpo @syarra

PS: We talked about this a long time ago. I'm in the process to add tests, tests and again tests. It would be easier for me to be able to generate tests for all cases (complex and real) at once. But we also can keep things like they are.

dpo commented 8 years ago

Certainly for real matrices, we should have A.conj == A and A.H == A.T, etc.

ghost commented 8 years ago

I knew it!

ghost commented 8 years ago

Done but need testing.

ghost commented 8 years ago

Testing done for all cases for the multiplications with a NumPy vector.