README.txt

"""
________________________________________
ROBOTRAINING PACKAGE

----- Requirements -----
written for Python 2.7 on Windows OS
using packages:
	numpy
	matplotlib
	cma
	neurolab
	sqlite3
	pyevolve
	
in order to create movies: FFmpeg must be installed	

----- Examples -----

Create a movie of a moving robot :
	from roboTraining.robot import *
	from roboTraining.simulate import *

	env = HardEnvironment() 
	morph = SpringMorphology(noNodes = 20,spring = 50, environment = env)
	control = SineControl(morph,amplitude = 0.3) 
	robot = Robot(morph, control)
	plotter = Plotter(movie = True, plot = True, movieName = "RobotMovie", plotCycle = 6, color = True);
	simulEnv = SimulationEnvironment(timeStep = 1.0/200, simulationLength = 12000, plot =  plotter)
	simul = VerletSimulation(simulEnv, robot)
	simul.runSimulation() # perform simulation;

Train the control signals of a robot to move by optimizing omega, phase, amplitude using GA:
	from roboTraining.robot import *
	from roboTraining.simulate import *
	from roboTraining.training import *

	env = HardEnvironment()
	morph = SpringMorphology(noNodes = 20 ,spring = 100, noNeighbours = 3,environment = env)
	control = SineControl(morph)
	robot = Robot(morph,control)

	plotter = Plotter(plotCycle = 20,plot=False);
	simulenv = SimulationEnvironment(1.0 / 200, simulationLength = 100, plot =plotter)
	
	trainscheme = TrainingScheme()
	trainscheme.createTrainVariable("omega", 0,10) # omega bounded between 0 and 10
	trainscheme.createTrainVariable("phase", 0, 2*np.pi)
	trainscheme.createTrainVariable("amplitude", 0, 0.25)

	saver = Save(None, 'RobotData', 'GeneticTraining') 
	train = GeneticTraining(trainscheme, robot, simulenv, saver = saver)
	param, score = train.run() # perform optimization
	train.save() # save all data and plots
________________________________________

"""