Error using cross validation with plsregress
显示 更早的评论
Hi everyone,
I am running a PLS regression and try to perform cross validation (k-fold) with matlab data.
load spectra
X = NIR;
y = octane;
[XL,yl,XS,YS,beta,PCTVAR,MSE] = plsregress(X,y,10,'CV',10);
Unfortunately, this error appear:
Could you explain to me how to solve it please?
Thank you in advance for your help.
Bruno
4 个评论
Adam Danz
2020-9-23
When I run the example code you shared, it runs smoothly in r2020b.
What release are you using?
You may have a file or function named "crossval" that is shadowing Matlab's crossval() function.
The line in the screenshot that indicates the error in that function contains variable names and commands that are not in the built-in crossval() function, at least not in r2020b. But that could also be explained by using an older release of Matlab.
To list functions with the name crossval,
which crossval -all
bruno ebel
2020-9-24
bruno ebel
2020-9-24
bruno ebel
2020-9-24
But errors are still here...
采纳的回答
Based on the results of
which crossval -all
a function with the same name from a 3rd party toolbox (pls toolbox) is shadowing Matlab's crossval function which is called by plsregress.
If you are not using the pls toolbox, remove it from your Matlab path using
rmpath('C:\Program Files\MATLAB\R2017a\toolbox\pls_toolbox_862')
% or
rmpath([matlabroot, '\toolbox\pls_toolbox_862'])
If you are using the pls toolbox, you can use Julian's answer explained here
It grabs the function handle from Matlab's crossval function and assigns it to a variable that you can use in addition to the crossval version in pls toolbox.
3 个评论
bruno ebel
2020-9-25
In the past when I faced the same problem I used Julian's answer explained here
I'll update my answer to include this.
bruno ebel
2020-9-29
更多回答（0 个）
类别
在 帮助中心 和 File Exchange 中查找有关 Gaussian Process Regression 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!
Translated by 
