84098

COMPARISON OF DIFFERENT PYTHON DEVELOPMENT LIBRARIES TO SOLVE FLUID FLOW PROBLEMS

Favoritar este trabalho

With the advance of programming languages, the availability of better
hardware, and the emergence of new libraries for numerical development, it
has been possible to use high level abstractions to prototype and solve
complex problems in many fields of knowledge. In the context of numerical
simulation using the Finite Volume Method, those new tools can be used, so
that the developer is able to make experiments with less effort. This work
compares several libraries from the Python programming language
(Pure Python implementation, Cython, NumPy, Numba, Scipy, petsc4py, and pure C/C++), aiming to solve fluid flow
problems and comparing execution performance from each tool