If you are looking for the code to find the errors you want to be here.
For the following, let W = | Um - Um+1 |

I. Max Error
The max error is the simplest error, where you just find the max value of the absolute difference between the two solutions: | W |

II. Average Error
The average error or L2 norm error is the square root of the integral of | W |^2.

III. Energy Error
The energy error is ( (5/3)^m * the sum over Vm of (W(y) - W(x))^2 )^(1/2) where y is a neighbor of x.

IV. A Posteriori Error
The posteriori error is basically L2 norm of -Lu+qu-f. The advantage here is that there is no reference to the theoretical solution, thereby allowing us to measure the error of any solution to any f and q functions. It should be noted that the slopes associated with the energy error tends to be the average of the slopes calculated from the average error and the posteriori error. This is because, generally speaking, each derivative you take makes the approximation worse, and the average error involves no derivatives, while the energy error involves the first derivative (squared), and the posteriori error operates on the level of the laplacian, which is like 2 derivatives, therby yielding the worst errors.