Error in using fitgmdist

2016 12 28
1 个回答
更新时间：2016 12 29
4 次查看（30 天）

0 个投票

Hi, I'm try to use fitgmdist to cluster high dimensional data with size of 96*879. Using
gmfit = fitgmdist(data,3,' CovarianceType','full','RegularizationValue',0.1);
Then error occurs:
X must have more rows than columns.
How can I fix this?
Thanks.

请先登录，再回答此问题。

回答（1 个）

Jorge
Jorge 2016-12-28

0 个投票

The number of colums in data (879) corresponds to the dimensionality of the data. The number of rows (96) is the number of observations.
You may need to transpose your data before using it.

3 个评论
显示 1更早的评论 隐藏 1更早的评论

Jorey
Jorey 2016-12-28
I think this can change the structure of the data for clustering, as using X is clustered according to rows, i.e., samples, but using X' is clustered according to variables. This can cause false conclusions, am I right?
Jorge
Jorge 2016-12-28
编辑：Jorge 2016-12-28
You are right. I'm assuming that your data is composed of 879 samples, each sample with a dimensionality of 96. If that is the case, then you should pass data' (transposed) instead of data. If that is not the case, then you can't use fitgmdist.

请先登录，再进行评论。

类别

在 帮助中心 和 File Exchange 中查找有关 Gaussian Mixture Models 的更多信息

标签

提问:

Jorey
Jorey
2016-12-28

评论：

Jorey
Jorey
2016-12-29

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by