Hi I was exploring the AMORE package, to check if I can implement and train networks with customised architectures. I am basically training a network which has 2 inputs, 2 neurons in the hidden layer and 2 output neurons. I aim to remove the connection between a neuron from the hidden layer and a neuron in the output layer. I am using the following code :- #initialising of the network net.custom <- newff(n.neurons=c(2,2,2), learning.rate.global=1e-2, momentum.global=0.5, error.criterium="LMS", Stao=NA, hidden.layer="tansig", output.layer="purelin", method="ADAPTgdwm") #Remodeling of the network to remove connection between neuron[1] and neuron [4] net.custom$neurons[[1]]$output.links <- c(3) net.custom$neurons[[1]]$output.aims <- c(1) net.custom$neurons[[4]]$weights <- net.sofm$neurons[[4]]$weights[1] net.custom$neurons[[4]]$input.links <- c(2) net.custom$neurons[[2]]$output.aims <- c(2,1) #Initializing matrices for training of the network P <- matrix(c(c(1,2,3,4),c(1,2,3,4)),nrow = 4, ncol=2) target <- P #Just to keep things simple for now. # We train the network according to P and target. result <- train(net.sofm, P, target, error.criterium="LMS", report=TRUE, show.step=100, n.shows=5 ) ## Running the above line gives me the following error Error in train.method(net, P, T, show.step) : INTEGER() can only be applied to a 'integer', not a 'double' I was unable to fix the error by using P <- as.interger (martix(....)). Apart from this we could not think of any other method to fix this problem. The sessionInfo is as follows :- R version 2.10.1 (2009-12-14) i386-pc-mingw32 locale: [1] LC_COLLATE=English_United States.1252 [2] LC_CTYPE=English_United States.1252 [3] LC_MONETARY=English_United States.1252 [4] LC_NUMERIC=C [5] LC_TIME=English_United States.1252 attached base packages: [1] stats graphics grDevices utils datasets methods base other attached packages: [1] AMORE_0.2-12 > I wonder changing the values of ouput.links, input.links etc is sufficient for changing the connections between the neurons. Is there any other package in R which allows me to design and train neural network. Thank You Suyog P -- Masters Student Department of Chemical and Biological Eng, University at Buffalo, Buffalo, NY . Ph: 801 814 3893 [[alternative HTML version deleted]]