OwlDE is a lightweight package for solving ordinary differential equations, extending Owl’s mathematical exploration capabilities and bringing it one step closer to general-purpose frameworks like NumPy+SciPy and Julia+JuliaDiffEq.
Built on top of Owl’s numerical library, OwlDE is designed with extensibility and ease of use in mind: it already includes bindings for SUNDIALS’ CVODE and ODEPACK, and native implementations of classic ODE solvers (e.g. Euler and Runge-Kutta, in both adaptive and fixed-step variants) and symplectic solvers (e.g. Leapfrog), and an external library implementing native contact geometry solvers.
The design and implementation, and the planning of the remaining work, have pointed out multiple limitations of Owl, ranging from the way it builds to the design and implementation of some of Owl’s functionalities, and have thus allowed us to clarify what still needs to be added to the Owl ecosystem before it could be truly considered a dynamic exploratory scientific computing tool.
Fri 23 AugDisplayed time zone: Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna change
10:30 - 12:00 | |||
10:30 25mTalk | OwlDE: making ODEs first-class Owl citizens OCaml Marcello Seri Bernoulli Institute for Mathematics, Computer Science and Artificial Intelligence, University of Groningen, Ta-Chu Kao Computational and Biological Learning Lab, Department of Engineering, University of Cambridge | ||
10:55 25mTalk | CausalRPC: traceable distributed computation OCaml Craig Ferguson Tarides | ||
11:20 25mTalk | Executing Owl Computation on GPU and TPU OCaml Jianxin Zhao University of Cambridge | ||