I was inspired by this and came up with another interesting, short solution using numpy's rot90: https://www.codewars.com/kata/reviews/5412ed87648162f1e60000c4/groups/6399b331f70c1e000184df8f

You can very easily turn it into pure python by replacing the last line with a function that goes through the return value in reverse order.

Thanks a lot for helping me in my first kata!

Ok, fixed! The single-digit example is there, it's the 9 (the stem is 0).

Mostly done, but what do you mean by more random tests? Simply repeat more random tests like the one that is already there?

Thanks a lot!! I fixed those two points.

What is happening inside the hypot?

Your solution does not always work; if you attempt a few times you'll see that it fails sometimes. The problem is that params.items() does not guarantee the order of the tuples, so sometimes you get the right order, sometimes you don't. One deterministic right solution is to use collections.OrderedDict, for example.

Great idea! Also, I like it how you unrolled the tuple, much more readable than mine.

You are correct, but in the end all it does is it skips one iteration (where you would have removed a and c). Not sure if it's an intended skip, but it's actually clever.

In command you could check if cmd not in self.cmd and set it to 0, then you don't need to duplicate the return statement. Or you could simply use a defaultdict(int)!