Python for Contests contents

Understanding Verdicts

What Accepted, Wrong answer, Runtime error and Time limit exceeded mean on PyInfo, and a checklist for each failure.

Beginner3 min readjudgedebuggingverdicts

Every submission ends with a verdict. Reading it correctly is half of debugging.

Verdict	Meaning
Accepted	every test passed
Wrong answer	the program finished, but its output differs from the expected one
Runtime error	the program crashed (exception or non-zero exit code)
Time limit exceeded	it ran longer than the limit on some test
Compile error	a syntax error: Python could not even parse the file

The judge runs the tests in order and stops at the first failure; the message tells you which test failed and, for Wrong answer, what was expected and what you printed.

Wrong answer

The logic is wrong, or the format is. Check in this order:

Format. Extra prompt text, wrong capitalization (Yes vs YES), extra spaces, a missing decimal precision. Compare the expected and obtained lines character by character.
The samples. Does your program give the sample output? If not, trace it by hand on the sample.
Edge cases. $n = 1$ , empty or single-character strings, all elements equal, the minimum and maximum values, zero, negative numbers.
Integer versus float division. Use // for integers; / always returns a float, which prints as 5.0.
Off-by-one errors. Is range(n) supposed to be range(n + 1)? Is a slice end exclusive?

python

assert 7 / 2 == 3.5 and 8 / 2 == 4.0     # a float, prints as 4.0
assert 8 // 2 == 4                        # an int, prints as 4
assert list(range(3)) == [0, 1, 2]        # the stop value is excluded

Runtime error

Read the last line of the message: it names the exception.

Message	Likely cause
`ValueError: invalid literal for int()`	you read a line that is not a single integer, or forgot `.split()`
`IndexError: list index out of range`	the list is shorter than you assumed; check empty input and loop bounds
`EOFError: EOF when reading a line`	you call `input()` more times than there are lines
`ZeroDivisionError`	a division or modulo by a value that can be zero
`RecursionError`	recursion too deep; convert to a loop or use an explicit stack
`KeyError`	use `dict.get` or `in` before reading a key

Time limit exceeded

The tests may be much larger than the samples. Work out the complexity of your solution and compare it with the limits from Performance. The usual culprits:

two nested loops over large data ( $O(n^2)$ ); look for a dictionary, a set, sorting, prefix sums or a better algorithm,
x in some_list inside a loop,
reading with input() or printing in a loop for hundreds of thousands of lines (Fast I/O),
an infinite loop because the loop variable is never updated.

Compile error

Python reports the line: a missing colon, unbalanced brackets, inconsistent indentation, or a match statement on an interpreter older than 3.10. Run the file locally first.

A debugging routine that works

Reproduce the failing test locally: save its input to a file and run python3 solution.py < input.txt.
Write a slow but obviously correct brute force solution.
Generate small random tests and compare both programs. The first difference is a minimal counter-example.
Print intermediate values to sys.stderr (print(x, file=sys.stderr)): the judge only looks at standard output, so debug prints there do not break the format.