Is it possible to get a solution faster than O(m*n) runtime with O(n) additional space or no?

Why.

thanks for bringing this up. actually a 0 dimensional matrix because it fails to contain any values. it is an empty shell; a random tests byproduct for frustration purposes. would help to see this specific empty matrix in standard tests.

:|

I didn't use any library (NumPy) and works well. However of course is more efficient using NumPy, but I think is more interesting try without.

Really nice little 7 kyu task

numpy or not numpy?

Description doesn't mention numpy. Sample tests use lists, not numpy arrays. Full tests use numpy. Reference sulution uses numpy, and doesn't pass sample tests!

Is this kata about numpy?

• yes -> numpy should compare somewhere in description or sample tests, we don't have a crystal ball
• no -> numpy should not be used in full tests

naptın mk

Sure, Got it! It was my mistake in interpreting the kata.

The error message is wrong, but the test is correct. build_one_two_three().next.next.next is None, and it has no data attribute.

Python: You have a mistake in the test case number 8 ---> test.assert_equals(build_one_two_three().next.next.next, None, "Third node should should have 3 as data.") as you are referencing to the memory address and not to the data itself, thus an issue.

I think it should be ----> test.assert_equals(build_one_two_three().next.next.next.data, None, "Third node should should have 3 as data."). The test makes the entire test attempt to fail.

Check your res2 variable you are indeed very close to the solution, just need to format it a little. :}

Example code block in Python has a typo: sum_diagonals[ should be sum_diagonals([

In your test cases:

matrix = [] ---> This is not 0x0 matrix, it should be [[]]
test.assert_equals(sum_diagonals(matrix), 0) ```

some solutions does not work until you format it to 0x0, the description is clear we work with a 2 dimensional array and not 1 dimension. Hence an issue.