Help with speeding up scatteredInterpolant, parallel processing toolbox applicable?

3 次查看（过去 30 天）

Brian 2014-11-25

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/164187-help-with-speeding-up-scatteredinterpolant-parallel-processing-toolbox-applicable

评论： Mike Jackson 2016-6-1

Hello, thanks for reading this,

I want to possibly speed up a process, but everything I've read about parfor, for instance, makes this seem not applicable.

What I have is a massive point cloud (55k points) with four column entries for the 55k points, three for x y and z coordinates and the last for the data I use for coloring.

To plot the data, I use scatteredInterpolant, then create a meshgrid of the interpolated data. I have a second question regarding this process, which I will not ask here, but I was wondering if this process itself can be done in parallel processing because it takes VERY long for decently high resolutions.

This is the code I used:

https://gist.github.com/anonymous/8b3eab89bc708155900c

I do have the parallel processing toolbox.

Any advice would be helpful.

Thanks!

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Mike Jackson 2016-6-1

I have exactly the same problem, it is something I understand as of early 2016 is not supported by Matlab. It's a big gap that an algorithm this easily parallelizable is not implemented with support from the parallel processing toolbox.

请先登录，再进行评论。

请先登录，再回答此问题。