Fix bug with creation of extensions of function fields #41096
Conversation
|
Documentation preview for this PR (built with commit 74deb42; changes) is ready! 🎉 |
| EXAMPLES:: | ||
| sage: K.<x> = FunctionField(QQ); R.<y> = K[] | ||
| sage: K.extension(y^5 - x^3 - 3*x + x*y) # needs sage.rings.function_field |
There was a problem hiding this comment.
Now, we can not use K.extension like this? I think it is better to preserve these tests
There was a problem hiding this comment.
I do not understand, isn't it exactly line 456 in function_field.py? I feel like these tests were run twice, but maybe there is something silly I am missing?
There was a problem hiding this comment.
I do not understand, isn't it exactly line 456 in
function_field.py? I feel like these tests were run twice, but maybe there is something silly I am missing?
The base field is QQ, a rational number field. But in line 456 examples, its base field is GF(2). This examples is used to detect the extension method on the function field based on a rational number field.
There was a problem hiding this comment.
I do not understand, isn't it exactly line 456 in
function_field.py? I feel like these tests were run twice, but maybe there is something silly I am missing?
I think you can add this into the line456 test.
There was a problem hiding this comment.
Maybe I need another cup of black tea, I agree that the tests I added (starting line 475) are on GF(2), but it seems to me that the tests between line 456 and 469 are precisely the ones I deleted, with QQ. I just ran diff on these lines and the only difference is lines 463/464 vs 454/455. Maybe we can keep the small difference? But I don't see the GF(2) thing?
|
OK. Thank you. |
sagemathgh-41096: Fix bug with creation of extensions of function fields When creating a function field extension, it is never checked that the polynomial defining the extension is defined over that function field. Hence, when calling `base_ring` one gets a bug because the base ring of the polynomial ring is returned, and not the function field which we used to create the extension. Fixes sagemath#41095. When working on this, I remarked that the `extension` method of `RationalFunctionField` was just a copy paste (even for doctests) of the one of `FunctionField`, which is its parent. So I also removed it. But I can put it back if someone explains me why this can sometimes be useful to do so. ### 📝 Checklist <!-- Put an `x` in all the boxes that apply. --> - [X] The title is concise and informative. - [X] The description explains in detail what this PR is about. - [X] I have linked a relevant issue or discussion. - [X] I have created tests covering the changes. - [X] I have updated the documentation and checked the documentation preview. URL: sagemath#41096 Reported by: Rubén Muñoz--Bertrand Reviewer(s): Chenxin Zhong, Rubén Muñoz--Bertrand
3 participants
When creating a function field extension, it is never checked that the polynomial defining the extension is defined over that function field. Hence, when calling
base_ringone gets a bug because the base ring of the polynomial ring is returned, and not the function field which we used to create the extension.Fixes #41095.
When working on this, I remarked that the
extensionmethod ofRationalFunctionFieldwas just a copy paste (even for doctests) of the one ofFunctionField, which is its parent. So I also removed it. But I can put it back if someone explains me why this can sometimes be useful to do so.📝 Checklist