How to do it...

The following is the code to draw a contour plot of loss function in the given regression problem:

1. Read the Loss, theta0, and theta1 values from saved files:

Loss = pd.read_excel('Loss.xlsx')
theta0_vals = pd.read_excel('theta0.xlsx')
theta1_vals = pd.read_excel('theta1.xlsx')

2. Specify the figure size:

fig = plt.figure(figsize=(12,8))

3. Create the mesh grid for the X and Y coordinates:

X, Y = np.meshgrid(theta0_vals, theta1_vals)

4. Draw the contour plot and label contours with their corresponding loss values:

CS = plt.contour(X, Y, Loss, np.logspace(-2,3,20), cmap=cm.coolwarm)
                 plt.clabel(CS, inline=1, fontsize=10)

5. Display the contour plot on the screen:

plt.show()