herelol.py

import numpy as np

# ** DATASET **
# https://archive.ics.uci.edu/ml/machine-learning-databases/servo/ 
# ** DATASET **


# sigmoid function
def nonlin(x,deriv=False):
    if(deriv==True):
        return x*(1-x)
    return 1/(1+np.exp(-x))

inputDat = []
outputDat = []
with open('servo.data.txt','r') as neural_dat:
    print "opened the thing"
    rowsRaw = neural_dat.read().split('\n')
    for rowRaw in rowsRaw:
        netInputAndOutput = rowRaw.split(' ')
        if len(netInputAndOutput) == 2:
            netInputValues = netInputAndOutput[0]
            netOutputValues = float(netInputAndOutput[1])/5
            colsRaw = np.array([nonlin((ord(x)-65)/5,False) for x in netInputValues.split(',')[0:-1]]) #theres a leftover string I think, hence -1
            inputDat.append(colsRaw)
            outputDat.append(netOutputValues)

row = 0
for inputSet in inputDat:
    print "{}".format(inputSet) + " -> " + "{}".format(outputDat[row])
    row = row + 1
print " - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - "


# input dataset
# X = np.array([  [1,0,1],
#                 [1,1,1],
#                 [0,0,1],
#                 [0,1,1]
#                  ])
#
# # output dataset
# y = np.array([[1,1,0,0]]).T

# seed random numbers to make calculation
# deterministic (just a good practice)
np.random.seed(1)

X = np.array(inputDat)
y = np.array([outputDat]).T

# initialize weights randomly with mean 0
syn0 = 2*np.random.random((4,1)) - 1

for iter in xrange(10000):
    # print iter
    # forward propagation
    l0 = X
    l1 = nonlin(np.dot(l0,syn0))

    # how much did we miss?
    l1_error = y - l1

    # multiply how much we missed by the
    # slope of the sigmoid at the values in l1
    l1_delta = l1_error * nonlin(l1,True)

    # update weights
    syn0 += np.dot(l0.T,l1_delta)

print "Output After Training:"
# print l1
print syn0
print "{}".format(X[5]) + " should -> " + "{}".format(y[5])
print "{}".format(X[8]) + " feed forward -> " + "{}".format(nonlin(np.dot(X[8],syn0))) # should be around [ 0.5         0.5         0.04742587  0.04742587] -> **0.0712502961**