Fix construction functor for p-adic relaxed types. #35441
Merged
Conversation
This file contains 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
2 tasks
|
It looks like the linting errors were here. The PR looks fine otherwise. |
|
Alright, it seems like you need to merge #35418 here as well. I'll give both tickets positive review once tests pass. |
|
How come this is not caught by |
On sage 9.8: The same thing works ok after this PR, I'll add running a doctest to run the testsuite as above. |
|
Unrelated (random) doctest failure, which I can reproduce locally with this particular random seed: |
|
@roed314 all tests pass now here and in #35442. Except for:
For the failure, I'll have a look and submit a PR if it's trivial to fix or open an issue otherwise. |
|
Is it normal that github mentions that the file |
It is deleted in #35418. It shows up here because of the merge. That and the change to The current PR only changes two lines of code, the rest is all adding a bunch of new tests that fail without this PR and pass with this PR. |
|
I see, thanks for the answer. |
xcaruso
approved these changes
Apr 6, 2023
4 tasks
The precision for relaxed p-adics has 3 components:
`(default_prec, halting_prec, secure)`
where the last one `secure` is a boolean defaulting to `False`.
However, the `construction()` method doesn't know about it:
```
sage: K = QpER(5, secure=True)
sage: K.construction(forbid_frac_field=True)
(Completion[5, prec=(20, 40)], Rational Field)
sage: R = ZpER(5, secure=True)
sage: R.construction()
(Completion[5, prec=(20, 40)], Integer Ring)
```
This has two undesired consequences for the `change()` method:
a. The `secure` attribute is not copied:
```
sage: K.is_secure()
True
sage: K.change().is_secure()
False
sage: R.is_secure()
True
sage: R.change().is_secure()
False
```
b. The `check=False` option is broken:
```
sage: K.change(check=False)
...
ValueError: not enough values to unpack (expected 3, got 2)
sage: R.change(check=False)
...
ValueError: not enough values to unpack (expected 3, got 2)
```
Fixing the `construction()` method solves both issues.
After this commit:
```
sage: K = QpER(5, secure=True)
sage: K.construction(forbid_frac_field=True)
(Completion[5, prec=(20, 40, True)], Rational Field)
sage: K.change().is_secure()
True
sage: K.change(check=False)
5-adic Field handled with relaxed arithmetics
sage: K.change(check=False).is_secure()
True
sage: R = ZpER(5, secure=True)
sage: R.construction()
(Completion[5, prec=(20, 40, True)], Integer Ring)
sage: R.change().is_secure()
True
sage: R.change(check=False)
5-adic Ring handled with relaxed arithmetics
sage: R.change(check=False).is_secure()
True
```
tornaria
force-pushed
the
Qp_relaxed_construction
branch
from
d9dbfe5 to
1d0a04a
Compare
|
Rebased to 10.0.beta8 |
|
Documentation preview for this PR is ready! 🎉 |
roed314
approved these changes
Apr 7, 2023
vbraun
pushed a commit
that referenced
this pull request
Apr 13, 2023
gh-35442: Make Qp.integer_ring() faster. ### 📚 Description The method pAdicGeneric.integer_ring() uses LocalGeneric.change() to turn a p-adic field into a p-adic ring. The latter calls a factory function which, by default, checks primality of p. However, when p came from a Qp this step is not necessary. We avoid it by adding `check=False` to the call to `LocalGeneric.change()` in `pAdicGeneric.integer_ring()`. This results in significant time savings for large primes, e.g. in the current test suite: Before this commit: ``` sage: R = Qp(next_prime(10^60)) sage: timeit('R.integer_ring()') 25 loops, best of 3: 22.2 ms per loop sage: %time TestSuite(R).run() CPU times: user 14.4 s, sys: 44 µs, total: 14.4 s Wall time: 14.4 s ``` After this commit: ``` sage: R = Qp(next_prime(10^60)) sage: timeit('R.integer_ring()') 625 loops, best of 3: 68 μs per loop sage: %time TestSuite(R).run() CPU times: user 714 ms, sys: 239 µs, total: 715 ms Wall time: 717 ms ``` Doctest of `padic_base_leaves.py` goes down from ~33 to ~5 seconds. Note that the `check=False` option for the `change()` method in relaxed type is broken, so this needs #35441. Other than that this is a one- liner. ### 📝 Checklist - [x] The title is concise, informative, and self-explanatory. - [x] The description explains in detail what this PR is about. ### ⌛ Dependencies - #35441 URL: #35442 Reported by: Gonzalo Tornaría Reviewer(s): roed314
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
The precision for relaxed p-adics has 3 components:
where the last one
secureis a boolean defaulting toFalse.However, the
construction()method doesn't know about it:This has two undesired consequences for the
change()method:a. The
secureattribute is not copied:b. The
check=Falseoption is broken:Fixing the
construction()method solves both issues.After this commit:
📝 Checklist