Closed adam-hartshorne closed 2 months ago
Hm. The dependence has always been there. You might not have encountered it because you haven't used the Runge-Kutta starter. That said, it should not exist. I will take care of this.
No I definitely using autodiff.taylor_mode_scan before.
No I definitely using autodiff.taylor_mode_scan before.
I thought so :)
The issue has been resolved, and a new release has been published. Upgrade your probdiffeq version, and you should be good to go. Thanks for reporting.
Thanks.
With v0.4.0 and the rejigging of imports, you have now made a dependence on diffeqzoo and diffrax.
taylor -> ode -> diffeqzoo.
from probdiffeq import ivpsolve, ivpsolvers, stats, taylor
taylor.py
from probdiffeq.backend import control_flow, functools, itertools, ode, tree_util
ode.py