Thanks a lot for your help! The workaround works and more importantly I
understand what happened now. Unfortunately now I get a massive stack trace
and a segmentation fault... I guess I'll take that over to the DASSL
developers.
On Wednesday, October 19, 2016 at 7:17:23 PM UTC+2, Steven G.
A workaround from pyjulia might be to create a Julia function that can wrap
a Julia Function around a Python function
fwrap = j.eval('f -> (args...) -> f(args...)')
and then do
j.dasslSolve(fwrap(f), 1.0, [0.0, 10.0])
Not super elegant, but occasionally inter-language calling requires this
it seems a near term work around could be had through PyCall ... ?
https://github.com/JuliaPy/PyCall.jl
On Wednesday, October 19, 2016 at 7:21:05 AM UTC-7, Steven G. Johnson wrote:
>
> It looks like this is a problem in the DASSL.jl package. They made the
> common mistake of
It looks like this is a problem in the DASSL.jl package. They made the
common mistake of over-specifying the types of their arguments, and in
particular they require you to pass a Julia ::Function argument for the
equations to be solved, rather than any callable object (here, the Python
any luck with replicating the example from the README at
https://github.com/pwl/DASSL.jl
The python script looks like this (sorry, thought attachments would be
inlined):
import julia
j = julia.Julia()
j.using("DASSL")
f = lambda t, x, dx: x - dx
r = j.dasslSolve(f, 1., [0., 10.])
