Asked by Chaou
on 28 Jul 2013

Hello. I'm trying to execute the simplest example for Pattern recognition network, given in this link http://www.mathworks.fr/fr/help/nnet/ref/patternnet.html for iris flowers classification.

[x,t] = iris_dataset; net = patternnet(10) net = train(net,x,t); view(net) y = net(x); perf = perform(net,t,y) classes = vec2ind(y)

**Error using struct
Conversion to struct from double is not possible.**

And it's the same error if I use the Neural Network interface. Can anyone help me, please? Thank you very much!

*No products are associated with this question.*

Answer by Greg Heath
on 28 Jul 2013

Accepted answer

I cut and pasted your code. It ran with no problem or messages.

Opportunities for recent engineering grads.

## 3 Comments

## Jan Simon (view profile)

Direct link to this comment:http://www.mathworks.com/matlabcentral/answers/83355#comment_161891

Please be so kind and post the complete error message. Most of all show us the line, which causes the error and explain the type of the used variables.

## ML guy (view profile)

Direct link to this comment:http://www.mathworks.com/matlabcentral/answers/83355#comment_220214

When I cut and paste his code (on a mac running R2014a and MacOS 10.9.3), I get this error:

## ML guy (view profile)

Direct link to this comment:http://www.mathworks.com/matlabcentral/answers/83355#comment_220217

I discovered that the cause of my replication of your error was the fact that I had redefined the 'softmax' function in my MATLAB workspace. There are a number of functions that would cause this error if you redefine them, such as: netsum purelin tansig dividerand mse adaptwb learngdm crossentropy softmax