@JZJacobZhu Hi Jacob. In your case, I would say that we can accept this output as '24'. The reason is, we should expect some error when doing floating number calculating, so we may not get exactly 24 even if the result IS 24 from the purely mathematical perspective. We use eps as an tolerance so we won't miss those results. Indeed you won't want a too large eps (say, 0.1 may not be good), but the choice of eps is totally empirical (I would say...)
@1120131418 No, it's not correct. When you say it returns True for that case, you mean in Python 2, right? But there it's wrong for other cases. You shouldn't even be testing it with Python 2, since it's clearly not meant for that.