Get predictions for the whole dataset instead of the first element (linear_regression)

[denkochev]

Hi @luisguiserrano, I was reading you book and in the chapter with linear regression I couldn't understand why we are using only first element instead of the whole dataset (like you described in the book):

def linear_regression(features, labels, learning_rate=0.01, epochs = 1000):
    ...
    predictions = features[0]*price_per_room+base_price
    ...

Then I figured out that there is a mistake here, I wanted to open pull request but someone got ahead of me (#59).