Yes, I am modifying a finite element code written in python. I would like to perform the operator assembly in Julia rather than python. This will require parsing the finite element data in numpy format. I would like to implement an iterative linear solver on the the global linear system, and such a solver would be much faster in Julia as well.
On Tuesday, October 18, 2016 at 4:56:14 PM UTC-4, Cedric St-Jean wrote: > > Could you provide a more concrete example of what you're trying to do? > > On Tuesday, October 18, 2016 at 4:42:51 PM UTC-4, Corbin Foucart wrote: >> >> Suppose that I have a large Python code; I would like to use Julia to >> operate on the python workspace variables at certain locations in the code. >> What occurs to me is to either: >> >> 1) write out all python workspace data to file, read data into julia, >> operate, save, read back into python (seems bad) >> 2) Call Julia code directly from python (I don't want to perform some >> trivial computation as in the examples I've found, I want to operate on the >> lists of numpy arrays) >> >> Is there an efficient or documented way to do this? >> >