fixunknows on nftool GUI matlab

I am using the nftool GUI to set up a regression neural network.

My database has various NaN (missing values). When I run the GUI, everything seems to go right. It gives me the performance and the regression graph.

I read that by line code you can add a processFcn named 'fixunknowns' to the network.

My question is: In the GUI, is the neural network making the fixunknows? How the GUI is procesing this NaN? When I generate the script, the fixunknows function does not appear.

I wonder if it is only possible to treat this NaN values on line code? Or... perhaps the GUI implements the fixnknowns automatically?

Thank you.

Answer 1

[Greg Heath]

I don't know for sure. However, I see no reason why you should not assume that functions called from command line and GUI use the same defaults for both.

Hope this helps.

Thank you for formally accepting my answer

[Luisa Linux]

If I assume that the functions called are the same, then, the GUI is not using fixunknows. Because, by default, this function is not used (see neural network toolbox documentation). So how the nftool is given me results? Can I trust its results? What is happening with that unknows?

Thanks for your help! I appreciate it!

[Greg Heath]

I was sure you were wrong about that so I checked the source code of feedforwardnet and , unlike the code of the obsolete 2010 newff, the default 'fixunknowns' does not appear in feedforwardnet.

So, the obvious first question is: What version of MATLAB are you using?

The second is ( to the MATLAB team responsible): Why has 'fixuknowns' been dropped as a default?

My advice is to check with the MATLAB team responsible to get a definitive answer.