animated_plot.py

"""
Simple animated demo for Online SVR. 2 features generated at 100 random samples, sent through a sine function, and used to predict and learn.

Author: Adam Werries, awerries@cmu.edu
"""
import sys
import numpy as np
from matplotlib import pyplot as plt
from matplotlib import animation

import online_svr

# Set up animated figure
fig = plt.figure()
ax = plt.axes(xlim=(0,1),ylim=(-1,1))
line1, = ax.plot([],[],'or',lw=2,label='Predicted')
line2, = ax.plot([],[],'ob',lw=2,label='Truth')
plt.legend()

def init():
    line1.set_data([],[])
    line2.set_data([],[])
    return line1,line2,

def animate(i):
    OSVR.learn(testSetX[i,:], testSetY[i])
    PredictedY = OSVR.predict(testSetX[0:i,:])
    line1.set_data(testSetX[0:i,0], PredictedY)
    line2.set_data(testSetX[0:i,0], testSetY[0:i])
    return line1,line2,


#testSetX = np.array([[0.1],[0.2],[0.3],[0.4],[0.5]])
testSetX = np.random.rand(100,2)
testSetY = np.sin(2*np.pi*testSetX[:,0] + testSetX[:,1])
OSVR = online_svr.OnlineSVR(numFeatures = testSetX.shape[1], C = 10, eps = 0.1, kernelParam = 30, bias = 0, debug = False)
anim = animation.FuncAnimation(fig,animate,init_func=init,interval=1,frames=range(testSetX.shape[0]),repeat=False)
plt.show()