<meta http-equiv="Content-Type" content="text/html; charset=utf-8"><div dir="ltr">We are currently testing several implementations of the Frechet distance based on <div><a href="https://towardsdatascience.com/fast-discrete-fr%C3%A9chet-distance-d6b422a8fb77" target="_blank">https://towardsdatascience.com/fast-discrete-fr%C3%A9chet-distance-d6b422a8fb77</a><br></div><div>which itself is based on</div><div><a href="https://dl.acm.org/doi/10.1145/3150919.3150924" target="_blank">https://dl.acm.org/doi/10.1145/3150919.3150924</a><br></div><div>Indeed, a typical "trip" of a ship, e.g., from the available of the Danish Maritime Authority</div><div><a href="https://www.dma.dk/SikkerhedTilSoes/Sejladsinformation/AIS/Sider/default.aspx" target="_blank">https://www.dma.dk/SikkerhedTilSoes/Sejladsinformation/AIS/Sider/default.aspx</a><br></div><div>contains 10K-30K timestamped points and thus we need an extremely efficient implementation.</div><div><br></div><div>The current experiments are being developed here</div><div><a href="https://github.com/estebanzimanyi/MobilityDB/blob/frechet/src/point/tpoint_discfrechet.c" target="_blank">https://github.com/estebanzimanyi/MobilityDB/blob/frechet/src/point/tpoint_discfrechet.c</a><br></div><div>but this is still far from complete. We will benchmark the various alternatives and choose the right one(s) to make available in MobilityDB.</div><div><br></div><div>After that we will make the results available in the MobilityDB wiki.<br></div><div><br></div><div><br></div><div><br></div></div>