What the function "preparets" does in NARX neural network toolbox

I am using the NN timeseries App(NARX), I didn't understand the idea of using function "preparets" my input is 13*15 and my target is 13*1 but after running this line of code "[x,xi,ai,t] = preparets(net,X,{},T); " the number of rows will be reduced to 11 in input and target. I am wonder what's going on after applying this function

Thanks for your help

Answer 1

[Greg Heath]

The data in 2 rows are taken to provide the initial delay conditions,

Hope this helps.

Thank you for formally accepting my answer

Greg

[Sara]

Dear Greg,

Thanks for your answer. I already read some of your previous tutorials but still, I am confused about using the NARX. especially multi-steps-ahead prediction. I hope you do a favor and help me in this regard.

I want to use this algorithm for fault prognostic in a system. The input data is the data from run to failure of a system and the target is the health condition of my system at each stage. I've trained this data set and I've got good results in open-loop. however, my problem is if I have new data can this algorithm predict one step-ahead target? if yes how does it work. For instance if I have the input data of stage1 does this algorithm predict target stage 2?