Dear all, I have a quick question regarding what numerical integration method does DIPY's "LocalTracking" use to generate the Streamlines. Does it use Runge-Kutta integration? If not, may I ask why? Thank you and cheers, Joan Joan Falco-Roget PhD Student [cid:534717a6-66b0-4efa-9c6b-faff78cf2ce6] Centre for Computational Personalised Medicine International Research Foundation Czarnowiejska 36, building C5 30-054 Kraków Poland Phone number: +34 610 068 599 j.roget@sano.science<mailto:n.surname@sano.science> https://sano.science<http://www.sano.science/>
Hi Joan, Thank you for your question. We are using Euler integration. What we noticed since many years ago is that RK requires many more computational resources for small improvement in quality. Others may have different experiences/opinions. Nonetheless, it shouldn't be hard to add RK given the flexibility of the tracking API in DIPY which I think is quite modular. Best regards, Eleftherios On Mon, Jan 17, 2022 at 4:35 PM Joan Falco Roget <j.roget@sanoscience.org> wrote:
Hi Joan, Thank you for your question. We are using Euler integration. What we noticed since many years ago is that RK requires many more computational resources for small improvement in quality. Others may have different experiences/opinions. Nonetheless, it shouldn't be hard to add RK given the flexibility of the tracking API in DIPY which I think is quite modular. Best regards, Eleftherios On Mon, Jan 17, 2022 at 4:35 PM Joan Falco Roget <j.roget@sanoscience.org> wrote:
participants (2)
-
Eleftherios Garyfallidis -
Joan Falco Roget