Re: [eigen] PyEigen 0.1 released

[ Thread Index | Date Index | More Archives ]

Hi Gael,

nice job, I'm sure many users will be very happy to see this

Thanks :)

meaning the overhead of such a wrapper is largely dominated by calls
to operator new. So maybe the conclusions of that could be: - is
there a way to reduce such heap allocations ? and if not: - for the
user prefer "inplace" operators whenever it is possible - functions
like add(a,b,r) doing r = a+b would be ugly but much faster than r =
a+b, well, unless there is a way in python to map r = a+b to
add(a,b,r) ? - I'm only nitpicking ? ;)

Yes, I'm aware there might be overhead with creating Python & Eigen objects. I'm planning to see if there's room for optimization there in the future. Note that currently PyEigen does have inline operators too, eg. a *= b, maybe I should benchmark them separately.

I'll also have to see how the module will be used in practice to see if *= style inline operators are enough or if I should add "mul(a, b, r)" style methods. Mapping functions that way can't be done due to the nature of Python.

Thanks for the input in any case, it's appreciated!


Mail converted by MHonArc 2.6.19+