import numpy as np
from random import *
from pylab import plot, ylim, show
import matplotlib

unit_step = lambda x: 0 if x < 0 else 1

training_data = [ (np.array([0,0,1]), 0),
                  (np.array([0,1,1]), 1),
                  (np.array([1,0,1]), 1),
                  (np.array([1,1,1]), 1),
]

w = np.random.rand(3) #This may not work

errors = []

eta = 0.2

n = 100

for i in range(n):
    x, expected = choice(training_data) # Is this an iterator?
    result = np.dot(w,x)
    error = expected - unit_step(result)
    errors.append(error)
    w += eta * error * x

for x, _ in training_data: # The , _ will skip the second elem of a tuple.
    result = np.dot(x,w)
    print("{}: {} -> {}".format(x[:2],result, unit_step(result))) #This is a useful print function

ylim([-1,1])
plot(errors)
show()

#Is this updating?

#
# import numpy as np
#
# #In class assingment 9-5-17
# #Create a vector w such that the unbiased classification
# #function will separate the following data points:
# #Perform perceptron learning on the following dataset. the initial weight vector is initialized to zero
#
# #ANSWER will use classify function to return dot product of w x, must make sure it's a number
# def classify(w, b, x):
#     if w @ x + b >= 0:
#         return 1
#     else:
#         return -1
#
# #What does classify result mean, need a place to store it
# def update(w, b, x, y):
#     y_p = classify(w, b, x)
#     if y_p != y:
#         w += y_p * x
#         #bias gets added in the same way as weight, except we add 1 or subtract 1 // after this not in class
#         b += y_p
#     else:
#         w += y_p
#         b += y_p
#     return y_p
#
# #(<0.1, 0.9>, -1)
# #(<0.9, 0.2>, +1)
# #(<0.3, 0.6>, -1)
# #(<0.7, 0.3>, +1)