Plotting future values of a signal using FNN and State Space Reconstruction method

This function is a Matlab function I develop during my University research project.

This code could be able to calculate and plot the future values of an input signal (my case is an ECG signal) using the method of FNN in Sate Space Reconstruction technique for stochastic signal.
Let assume you have already determined the values of lag and dimension, now you want to predict N iterations ahead using parts of your signal as a training set and the other as a validating set, this code could can handle the task like so:
>> sig = load('testECG.txt'); % Raw data file, each numeric value in a row, no headers
>> dimension = 154; % Specify the dimension
>> delay = 1; % Specify the time lag
>> iterate = 1000; % Predict the next 1000 values
>> percent = 0.8 % Percentage for the training set
>> plot_forecast2 --> prediction line is red
Please notice that this code is just some function that I create for my personal purpose without any guaranteed accuracy. It is still a developing function and it has no purpose of commercialization.
If you are interested please help me improve the code.
Contact email: [e-mail]