diff: support for inf/-inf #150
Conversation
| @@ -264,10 +264,19 @@ def are_different(first, second, tolerance): | |||
| return False | |||
|
|
|||
| first_is_nan, second_is_nan = bool(first != first), bool(second != second) | |||
| first_is_inf, second_is_inf = first == float('inf'), second == float('inf') | |||
| first_is_ninf = first == float('-inf') | |||
| second_is_ninf = second == float('-inf') | |||
|
|
|||
| if first_is_nan or second_is_nan: | |||
There was a problem hiding this comment.
on a more mathematical level, should two nan or two inf or two -inf be considered the same? It is improbable that two inf are the same when you are running a diff. If we agree that they should be considered the same, then this all looks great!
There was a problem hiding this comment.
I asked myself the same question while doing this PR and I found links like this stating that inf==inf and -inf==-inf .
On the contrary, it seems that they agree that Nan!=Nan , but that's outside the scope of this PR.
Let me know if you prefer not considering two inf the same, I can easily adapt tests and code.
There was a problem hiding this comment.
I think following convention is preferred over mathematical precision, so this sounds good to me 👍
Closes #146