Normal, boundary and invalid data; trace and fix. Learn through short examples, practical tasks and targeted questions.
Testing checks whether a program behaves as intended for normal, boundary and invalid inputs. A trace table records variable values as the algorithm runs, making errors easier to isolate. Change one part at a time and rerun the same test.
Watch out: A program working for one example does not prove it is correct. Include boundary values and unexpected input in the test plan.
Normal, boundary and invalid data; trace and fix. Study the code and predict what it does before running it. Change a value or string and observe how the output changes.
def discount(price):
if price >= 20:
return price * 0.9
return price
for value in (19.99, 20, 20.01):
print(value, discount(value))Create a function that accepts marks from 0 to 100, then test -1, 0, 50, 100 and 101. Explain which are invalid, normal or boundary cases.
Plan your inputs, processing and outputs. Write the code in Mu, IDLE or PyCharm; test at least three cases.
def valid_mark(mark):
return 0 <= mark <= 100
for mark in (-1, 0, 50, 100, 101):
print(mark, valid_mark(mark))Identify the problem and explain how to fix it before revealing the answer.
def average(a, b):
return a + b / 2def average(a, b):
return (a + b) / 2Compare the two snippets. Identify the changed line, run the corrected version and explain the effect.
Extension: Improve your solution to deal with unexpected input, add a second feature or display results more clearly.
Reflect: What was the hardest part? What would you test if someone else used your program?
Want extra practice? Try interactive Python tracing and debugging questions →