ROC on LM backrpopagation

Hello

when I run code below My ROC

as on example from

https://uk.mathworks.com/help/deeplearning/ref/plotroc.html

is single line as on screen

simpleclusterInputs = TrainSet';
simpleclusterTargets = TargetSet';

%load simplecluster_dataset
net = patternnet(20);
net = train(net,simpleclusterInputs,simpleclusterTargets);
simpleclusterOutputs = sim(net,simpleclusterInputs);
plotroc(simpleclusterTargets,simpleclusterOutputs)

How i can change input data to get correct output ? for each sclass as in sample