make assert_equal an example how to partial torch.testing.assert_close
#58918
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Instead of a distinct `torch.testing.assert_close` and `torch.testing.assert_equal`, this makes `torch.testing.assert_equal` a special case of `torch.testing.assert_close` for `rtol=atol=0`. In this case the closeness definition `abs(actual - expected) <= atol + rtol * abs(expected)` boils down to `abs(actual - expected) <= 0`. Since `abs(x)` can never be `<0`, this is equivalent to `abs(a - b) == 0` and this again boils down to `a == b`. This makes maintaing the module a lot easier, because we don't need to keep two functions in sync. [ghstack-poisoned]
This was referenced May 25, 2021
💊 CI failures summary and remediationsAs of commit 0f9e7b8 (more details on the Dr. CI page and at hud.pytorch.org/pr/58918): 💚 💚 Looks good so far! There are no failures yet. 💚 💚 This comment was automatically generated by Dr. CI (expand for details).Follow this link to opt-out of these comments for your Pull Requests.Please report bugs/suggestions to the (internal) Dr. CI Users group. |
rgommers
added
the
module: testing
|
Instead of defining |
|
Is there logic like pytorch/torch/testing/_core.py Line 107 in 3607478 somewhere, where this will short-circuit to a bitwise equivalence test? |
|
Lint failures are real |
Nope, in the current state we will always use We could also use an equal check if |
Instead of a distinct `torch.testing.assert_close` and `torch.testing.assert_equal`, this makes `torch.testing.assert_equal` a special case of `torch.testing.assert_close` for `rtol=atol=0`. In this case the closeness definition `abs(actual - expected) <= atol + rtol * abs(expected)` boils down to `abs(actual - expected) <= 0`. Since `abs(x)` can never be `<0`, this is equivalent to `abs(a - b) == 0` and this again boils down to `a == b`. This makes maintaing the module a lot easier, because we don't need to keep two functions in sync. [ghstack-poisoned]
….assert_close`" ~Instead of a distinct `torch.testing.assert_close` and `torch.testing.assert_equal`, this makes `torch.testing.assert_equal` a special case of `torch.testing.assert_close` for `rtol=atol=0`. In this case the closeness definition `abs(actual - expected) <= atol + rtol * abs(expected)` boils down to `abs(actual - expected) <= 0`. Since `abs(x)` can never be `<0`, this is equivalent to `abs(a - b) == 0` and this again boils down to `a == b`.~ Following #58918 (comment) and some offline discussions, we opted to use `assert_equal` as an example how to `partial` it. This makes maintaing the module a lot easier, because we don't need to keep two functions in sync. [ghstack-poisoned]
pmeier
changed the title
torch.testing.assert_close
pmeier
added a commit
that referenced
this pull request
Jun 17, 2021
…l `torch.testing.assert_close`" ~Instead of a distinct `torch.testing.assert_close` and `torch.testing.assert_equal`, this makes `torch.testing.assert_equal` a special case of `torch.testing.assert_close` for `rtol=atol=0`. In this case the closeness definition `abs(actual - expected) <= atol + rtol * abs(expected)` boils down to `abs(actual - expected) <= 0`. Since `abs(x)` can never be `<0`, this is equivalent to `abs(a - b) == 0` and this again boils down to `a == b`.~ Following #58918 (comment) and some offline discussions, we opted to use `assert_equal` as an example how to `partial` it. This makes maintaing the module a lot easier, because we don't need to keep two functions in sync. [ghstack-poisoned]
pmeier
commented
Jun 17, 2021
Comment on lines
648
to
652
| AssertionError: Tensors are not close! | ||
| Mismatched elements: 1 / 1 (100.0%) | ||
| Greatest absolute difference: 8.999999703829253e-10 at 0 (up to 0 allowed) | ||
| Greatest relative difference: 8.999999583666371 at 0 (up to 0 allowed) |
There was a problem hiding this comment.
This will make much more sense after #60091 from later in this stack is in. The error will read:
AssertionError: Scalars are not equal!
Absolute difference: 8.999999703829253e-10
Relative difference: 8.999999583666371
….assert_close`" ~Instead of a distinct `torch.testing.assert_close` and `torch.testing.assert_equal`, this makes `torch.testing.assert_equal` a special case of `torch.testing.assert_close` for `rtol=atol=0`. In this case the closeness definition `abs(actual - expected) <= atol + rtol * abs(expected)` boils down to `abs(actual - expected) <= 0`. Since `abs(x)` can never be `<0`, this is equivalent to `abs(a - b) == 0` and this again boils down to `a == b`.~ Following #58918 (comment) and some offline discussions, we opted to use `assert_equal` as an example how to `partial` it. This makes maintaing the module a lot easier, because we don't need to keep two functions in sync. [ghstack-poisoned]
pmeier
added a commit
that referenced
this pull request
Jun 18, 2021
…l `torch.testing.assert_close`" ~Instead of a distinct `torch.testing.assert_close` and `torch.testing.assert_equal`, this makes `torch.testing.assert_equal` a special case of `torch.testing.assert_close` for `rtol=atol=0`. In this case the closeness definition `abs(actual - expected) <= atol + rtol * abs(expected)` boils down to `abs(actual - expected) <= 0`. Since `abs(x)` can never be `<0`, this is equivalent to `abs(a - b) == 0` and this again boils down to `a == b`.~ Following #58918 (comment) and some offline discussions, we opted to use `assert_equal` as an example how to `partial` it. This makes maintaing the module a lot easier, because we don't need to keep two functions in sync. [ghstack-poisoned]
|
@mruberry has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Stack from ghstack:
msgintorch.testing.assert_close#60254 tests for diagnostics in callablemsgintorch.testing.assert_closetorch.testing.assert_close#60163 update docstring examples oftorch.testing.assert_closetorch.testing.assert_closein case of mismatching values #60091 Improve error messages oftorch.testing.assert_closein case of mismatching valuestorch.testing.assert_close#58918 make assert_equal an example how to partialtorch.testing.assert_closeInstead of a distincttorch.testing.assert_closeandtorch.testing.assert_equal, this makestorch.testing.assert_equala special case oftorch.testing.assert_closeforrtol=atol=0. In this case the closeness definitionabs(actual - expected) <= atol + rtol * abs(expected)boils down toabs(actual - expected) <= 0. Sinceabs(x)can never be<0, this is equivalent toabs(a - b) == 0and this again boils down toa == b.Following #58918 (comment) and some offline discussions, we opted to use
assert_equalas an example how topartialit.This makes maintaing the module a lot easier, because we don't need to keep two functions in sync.
Differential Revision: D29259404