Clustering with big data

Question

Filipa Cardoso 2018-7-17

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/410738-clustering-with-big-data

评论： Filipa Cardoso 2018-8-14

I have a vector with 5 million distances [5000000*1] obtained by using the pdist function and the goal is to do hierarchical cluster analysis. However when i use the linkage function i get the following error:

Error using linkage (line 240) Requested 1x12081952473360 (90017.6GB) array exceeds maximum array size preference. Creation of arrays greater than this limit may take a long time and cause MATLAB to become unresponsive. See array size limit or preference panel for more information.

I would like to know how to avoid this error and if that's not possible, which function I should use to make the cluster analysis.

Thanks in advance.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Bernhard Suhm 2018-8-6

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/410738-clustering-with-big-data#answer_331821

Maybe you are using the LINKAGE function incorrectly. You already has the pairwise distances, but for some reason the LINKAGE function thinks that you are passing the samples (X), not the distances (Y). LINKAGE should not have a trouble linking 5,000,000 distances, probably it is a matter of just putting the distances in the right form.