Perceptron neural network in python from scratch - Under 10 lines of code - works for and gate , or gate

perceptron.py

#Read this first - https://jontysinai.github.io/jekyll/update/2017/09/24/the-mcp-neuron.html
#Set The inputs and the desired Outputs for and gate.
x1,x2,y = [0,0,1,1],[0,1,0,1], [0,0,0,1]
#Set Random values for weights w1,w2 and for the bias b. initialize all to 0
w1,w2,b = 0,0,0
#try 100 iterations to see if the perceptron can guess that we need a fucking and gate.
for _ in range(100):
    #iterate for each input:
    for i in [0,1,2,3]:
        #just guess a y value based on our random weights and bias
        y_guess = w1*x1[i]+w2*x2[i]+b
        # convert y_guess to 0 or 1
        y_guess =  1 if y_guess > 0 else 0
        #for the guessed y => calculate the error/difference from actual y value
        error = y[i] - y_guess
        #now tune/tweak/adjust w1 to reduce error in next iteration
        w1 = w1 + 0.01*error*x1[i]
        #now tune/tweak/adjust w2 to reduce error in next iteration
        w2 = w2 + 0.01*error*x2[i]
        #now tune/tweak/adjust bias to reduce error in next iteration
        b = b + 0.01*error
#voila after 100 iterations , the neural network has figured out that the weights , bias to make an and gate for us :        
print (w1*0+w2*0+b)>0
print (w1*0+w2*1+b)>0
print (w1*1+w2*0+b)>0
print (w1*1+w2*1+b)>0    
#please follow me on twitter - @mrtechmaker
# print w1
# print w2
# print b   
#inspiration =>  https://medium.com/@thomascountz/19-line-line-by-line-python-perceptron-b6f113b161f3