How can I do a 50-50 split on data to obtain train and test datasets such that no value is common to both sets?

I am new to matlab and I can't find a function to do this.

Answer 1

[Walter Roberson]

http://www.mathworks.com/help/nnet/ug/divide-data-for-optimal-neural-network-training.html

"The divide function is accessed automatically whenever the network is trained, and is used to divide the data into training, validation and testing subsets. If net.divideFcn is set to 'dividerand' (the default), then the data is randomly divided into the three subsets using the division parameters net.divideParam.trainRatio, net.divideParam.valRatio, and net.divideParam.testRatio. The fraction of data that is placed in the training set is trainRatio/(trainRatio+valRatio+testRatio), with a similar formula for the other two sets. The default ratios for training, testing and validation are 0.7, 0.15 and 0.15, respectively."

net.divideParam.trainRatio = 0.5;
net.divideParam.testRatio = 0.5;
net.divideParam.valRatio = 0;

[Greg Heath]

Since you are new at this my advice is to begin using all of the defaults. Short examples are given in the documentation. See

REGRESSION/CURVEFITTING

 help fitnet
 doc fitnet

CLASSIFICATION/PATTERNRECOGNITION

 help patternnet
 doc patternet

For zillions of examples search in both the NEWSGROUP and ANSWERS using the search words

 greg fitnet
 greg patternnet

Hope this helps.

Greg