import numpy as np
from nn_mlp import BackPropogationNetwork,npA
if __name__ == '__main__':
import matplotlib
import matplotlib.pyplot as plt
print "Learning the sin function"
lFuncs = [None, BackPropogationNetwork.sgm, BackPropogationNetwork.linear]
bpn = BackPropogationNetwork( (1,4,1), lFuncs, 0.4)
samples = np.zeros(20, dtype=[('x', float, 1), ('y', float, 1)])
samples['x'] = np.linspace(0,1,20)
samples['y'] = np.sin(samples['x']*np.pi)
input = np.array([samples['x']]).T
target = np.array([samples['y']]).T
max = 100000
#lnErr = 1e-5
lnErr = 1e-2
for i in range(max+1):
err = 0.0;
for ii in range(len(input)):
err += bpn.trainEpoch(npA(input[ii]),npA(target[ii]))
if i%2500 == 0:
#if i%20 == 0:
print "iteration {0}\tError : {1:0.6f}".format(i,err)
#bpn.NN_cout("Train_"+str(i))
if err <= lnErr:
x.append(i*width+xmin)
return tuple(x)
def hist1dDraw(v1,bins):
P.hist(v1,bins,histtype='step')
#P.hist(vv, bins, normed=1, histtype='step', cumulative=True)
P.plot()
P.show()
if __name__ == '__main__':
import matplotlib
import matplotlib.pyplot as plt
lFuncs = [None, BackPropogationNetwork.sgm, BackPropogationNetwork.linear]
bpn = BackPropogationNetwork( (2,4,1), lFuncs, 0.4)
samples = np.zeros(50, dtype=[('x', float, 1),('x2', float, 1), ('y', float, 1)])
samples['x'] = np.random.normal(40., 5., 50)
samples['x2'] = np.random.normal(40., 5., 50)
samples['y'] = np.ones(50)
samples2 = np.zeros(50, dtype=[('x', float, 1),('x2', float, 1), ('y', float, 1)])
samples2['x'] = np.random.normal(40., 50., 50)
samples2['x2'] = np.random.normal(40., 50., 50)
samples2['y'] = np.zeros(50)
#samples['x'] += generate1dvalues(50,10,20)
#samples['y'] += np.zeros(50)