Efficient 1D Heat Equation Solver
Sandy Herho, Siti Kaban, Dasapta Erwin Irawan, Rubiyanto Kapid · 2023
This paper presents a Numba-based solver for the 1D Heat Equation, seamlessly blending Python's readability with Numba's dynamic Just-In-Time (JIT) compilation.The explicit method exhibits a notable runtime reduction from 8.324 s to 4.035 s, while the implicit method sees a more pronounced improvement, decreasing from 9.970 s to 1.195 s.Statistical tests confirm the statistical significance of these efficiency gains.Future research directions include extending the solver to multidimensional heat equations, exploring advanced parallelization techniques, and implementing dynamic parameter optimization strategies.Collaboration with domain experts for real-world applications is also envisioned to validate the solver's performance and impact.In summary, the symbiosis of Python and Numba in crafting an optimized 1D Heat Equation solver marks a pivotal advancement in efficient numerical solutions.This research holds promise for diverse scientific applications, ushering in a new era of computational efficiency.