Building Machine Learning Systems with Python - Third Edition

Starting with a simple straight line

Let's assume for a second that the underlying model is a straight line. Then the challenge is how to best put that line into the chart so that it results in the smallest approximation error. SciPy's polyfit() function does exactly that. Given data x and y and the desired order of the polynomial (a straight line has an order of 1), it finds the model function that minimizes the error function defined earlier:

fp1 = np.polyfit(x, y, 1)

The polyfit() function returns the parameters of the fitted Model function, fp1:

>>> print("Model parameters: %s" % fp1)Model parameters: [ 2.59619213 989.02487106]

This means the best straight line fit is the following function:

f(x) = 2.59619213 * x + 989.02487106

We then ...