Training NARNET with GA

Following on from a previous question could somebody help me generalise the solution to a time delay network.

- How can I use the Genetic Algorithm (GA) to train a Neural Network in Neural Network Toolbox 6.0.3 (R2009b)?

http://www.mathworks.com/matlabcentral/answers/100323

I tried to modify the script by changing the network architecture. But looks like I need to pass additional arguments to the GA function. I may also need to modify the anonymous function and the GA ref to this function

h = @(x) mse_test(x, net, inputs, targets);

Looking to adapt this method to get the GA to modifiy, nodes in the hidden layer, time delay and the weights. Looking to use FTDNN, timedelaynet, and Narnet architectures.

I may want to modify all three of these features at the same time or just the weights.