graphs: make init_short_digraph always sort neighbors but without the extra log complexity #38427

cyrilbouvier · 2024-07-26T14:43:48Z

This PR improve the init_short_digraph function that is used to initialize StaticSparseCGraph (used for immutable Graph and DiGraph).

Before, a boolean parameter sort_neighbors was used to specify if we wanted to sort the neighbors or not. It implied an extra log in the complexity (as qsort was called).
With this PR, the neighbors are always sorted at no extra cost. It is done by appending to the neighbors list the vertices in the correct order so the call to qsort is not needed anymore.

This PR partly reverts #38198 and mostly reverts #37662

📝 Checklist

The title is concise and informative.
The description explains in detail what this PR is about.
I have linked a relevant issue or discussion.
I have created tests covering the changes.
I have updated the documentation and checked the documentation preview.

⌛ Dependencies

…mplexity

cyrilbouvier · 2024-07-26T14:54:58Z

Setting this PR to a draft: the doctests that I added to init_short_digraph requires that the vertices are not sorted before being passed to init_short_digraph as it is currently done in __init__ of StaticSparseBackend.
I have a local commit that remove the vertices.sort() in StaticSparseBackend but it breaks one test in connectivity.pyx and I am not able to figure out why.

I have to:

either find another way to indirectly test init_short_digraph
or fix the doctests in connectivity.pyx when vertices are not sorted in StaticSparseBackend.

Note that the sorting of vertices in StaticSparseBackend probably invalidates most complexity claims of algorithms that rely on building Graph or DiGraph with immutable=True

github-actions · 2024-07-26T15:26:17Z

Documentation preview for this PR (built with commit 8627957; changes) is ready! 🎉
This preview will update shortly after each push to this PR.

cyrilbouvier · 2024-08-19T10:07:53Z

See issue #38527

cyrilbouvier · 2024-08-21T07:40:38Z

Failing doctest in graphs/connectivity.pyx will pass once PR #38535 is merged

…tic graphs

cyrilbouvier · 2024-08-22T13:00:48Z

In 71113cd I start to fix some tests whose output where changed by bd63dd7 (where sort is remove from StaticSparseBackend).

First it is not an easy task as it touches lots of different fields (LieAlgebra, Poset, ...) that I am not familar with. But from what I understand, static graphs are used in these structures (for example the hasse_diagram attribute of Poset is a static digraph) and are the reason for the changes in the output.
But maybe people more familiar with these objects should check my changes carefully.

Second, I hate tests whose output depend on internal representation of objects... 😠 (for example when the output is a list where the order does not matter but depends on the internal representation of the classes Graph and DiGraph). When possible I added sorted to avoid this problem. When it was not possible, I just changed the expected line of the doctests.

Special mention for the doctest of the method degree_on_basis in src/sage/algebras/lie_algebras/bgg_dual_module.py that never calls this method (so that is not a test for this method) but calls another method with the same name in src/sage/algebras/lie_algebras/verma_module.py. But I do not know what to do with this information.

…tic graphs

cyrilbouvier · 2024-08-22T14:59:54Z

I fixed all tests except the two from graphs/connectivity.pyx which are fixed by PR #38535

This PR can now be reviewed.

dcoudert

This is a smart improvement. Thanks. I have only minor comments.

The tests in connectivity.pyx should be fixed since #38535 as been merged.

I agree with you that doctests should not depend on the internal representation of objects. We reduced the number of such doctests, but we still have some.

src/sage/algebras/lie_algebras/bgg_dual_module.py

src/sage/categories/regular_crystals.py

src/sage/graphs/base/static_sparse_graph.pyx

src/sage/schemes/elliptic_curves/ell_field.py

cyrilbouvier · 2024-09-04T17:20:39Z

Changes are done.
I also merge the latest develop branch. It should fix doctests in graphs/connectivity.pyx.

dcoudert

This is a very nice improvement. Thank you.

LGTM.

…rs but without the extra log complexity       This PR improve the `init_short_digraph` function that is used to initialize `StaticSparseCGraph` (used for immutable `Graph` and `DiGraph`). Before, a boolean parameter `sort_neighbors` was used to specify if we wanted to sort the neighbors or not. It implied an extra `log` in the complexity (as `qsort` was called). With this PR, the neighbors are always sorted at no extra cost. It is done by appending to the neighbors list the vertices in the correct order so the call to `qsort` is not needed anymore. This PR partly reverts sagemath#38198 and mostly reverts sagemath#37662 ### 📝 Checklist  - [x] The title is concise and informative. - [x] The description explains in detail what this PR is about. - [ ] I have linked a relevant issue or discussion. - [x] I have created tests covering the changes. - [ ] I have updated the documentation and checked the documentation preview. ### ⌛ Dependencies    URL: sagemath#38427 Reported by: cyrilbouvier Reviewer(s): cyrilbouvier, David Coudert

vbraun · 2024-09-08T08:55:24Z

On 32-bit I'm getting

**********************************************************************
File "src/sage/algebras/lie_algebras/bgg_dual_module.py", line 204, in sage.algebras.lie_algebras.bgg_dual_module.BGGDualModule.degree_on_basis
Failed example:
    elt = Mc.an_element(); elt
Expected:
    f[-alpha[2]]^2*f[-alpha[5]]^2*f[-alpha[3]]^3*v[Lambda[1]
     + Lambda[4] - 1/3*Lambda[5]]^* + 2*f[-alpha[2]]*v[Lambda[1]
     + Lambda[4] - 1/3*Lambda[5]]^* + 3*f[-alpha[5]]*v[Lambda[1]
     + Lambda[4] - 1/3*Lambda[5]]^* + v[Lambda[1] + Lambda[4]
     - 1/3*Lambda[5]]^*
Got:
    f[-alpha[1]]^2*f[-alpha[3]]^2*f[-alpha[2]]^3*v[Lambda[1] + Lambda[4] - 1/3*Lambda[5]]^* + 2*f[-alpha[1]]*v[Lambda[1] + Lambda[4] - 1/3*Lambda[5]]^* + 3*f[-alpha[3]]*v[Lambda[1] + Lambda[4] - 1/3*Lambda[5]]^* + v[Lambda[1] + Lambda[4] - 1/3*Lambda[5]]^*
**********************************************************************
File "src/sage/algebras/lie_algebras/bgg_dual_module.py", line 210, in sage.algebras.lie_algebras.bgg_dual_module.BGGDualModule.degree_on_basis
Failed example:
    [M.degree_on_basis(m) for m in elt.support()]
Expected:
    [3*Lambda[1] - Lambda[2] - 2*Lambda[3] + 4*Lambda[4] - 4/3*Lambda[5],
     Lambda[1] + Lambda[4] - 1/3*Lambda[5],
     2*Lambda[1] - 2*Lambda[2] + Lambda[3] + Lambda[4] - 1/3*Lambda[5],
     Lambda[1] + Lambda[3] + Lambda[4] - 7/3*Lambda[5]]
Got:
    [-2*Lambda[2] - Lambda[3] + 3*Lambda[4] + 5/3*Lambda[5],
     Lambda[1] + Lambda[4] - 1/3*Lambda[5],
     -Lambda[1] + Lambda[2] + Lambda[4] - 1/3*Lambda[5],
     Lambda[1] + Lambda[2] - 2*Lambda[3] + 2*Lambda[4] + 2/3*Lambda[5]]
**********************************************************************
File "src/sage/algebras/lie_algebras/bgg_dual_module.py", line 612, in sage.algebras.lie_algebras.bgg_dual_module.SimpleModuleIndices.__contains__
Failed example:
    gens = list(I.gens()); gens
Expected:
    [f[-alpha[2]],
     f[-alpha[1]],
     f[-alpha[1] - alpha[2]],
     f[-2*alpha[1] - alpha[2]],
     f[-3*alpha[1] - alpha[2]],
     f[-3*alpha[1] - 2*alpha[2]]]
Got:
    [f[-alpha[1]],
     f[-alpha[2]],
     f[-alpha[1] - alpha[2]],
     f[-2*alpha[1] - alpha[2]],
     f[-3*alpha[1] - alpha[2]],
     f[-3*alpha[1] - 2*alpha[2]]]
**********************************************************************
File "src/sage/algebras/lie_algebras/bgg_dual_module.py", line 619, in sage.algebras.lie_algebras.bgg_dual_module.SimpleModuleIndices.__contains__
Failed example:
    gens[1] in I
Expected:
    True
Got:
    False
**********************************************************************
File "src/sage/algebras/lie_algebras/bgg_dual_module.py", line 900, in sage.algebras.lie_algebras.bgg_dual_module.SimpleModule.lift
Failed example:
    [L.lift(b) for b in L.basis()]  # long time
Expected:
    [v[Lambda[1]]^*,
     f[-alpha[1]]*v[Lambda[1]]^*,
     f[-alpha[2]]*f[-alpha[1]]*v[Lambda[1]]^* - f[-alpha[1] - alpha[2]]*v[Lambda[1]]^*,
     f[-alpha[1]]*f[-alpha[1] - alpha[2]]*v[Lambda[1]]^*
      + f[-2*alpha[1] - alpha[2]]*v[Lambda[1]]^*,
     f[-alpha[1]]^2*f[-alpha[1] - alpha[2]]*v[Lambda[1]]^*
      + f[-alpha[1]]*f[-2*alpha[1] - alpha[2]]*v[Lambda[1]]^*
      + 1/2*f[-3*alpha[1] - alpha[2]]*v[Lambda[1]]^*,
     f[-alpha[2]]*f[-alpha[1]]^2*f[-alpha[1] - alpha[2]]*v[Lambda[1]]^*
      + f[-alpha[2]]*f[-alpha[1]]*f[-2*alpha[1] - alpha[2]]*v[Lambda[1]]^*
      + 1/2*f[-alpha[2]]*f[-3*alpha[1] - alpha[2]]*v[Lambda[1]]^*
      - f[-alpha[1] - alpha[2]]*f[-2*alpha[1] - alpha[2]]*v[Lambda[1]]^*
      + 1/2*f[-3*alpha[1] - 2*alpha[2]]*v[Lambda[1]]^*,
     f[-alpha[1]]*f[-alpha[1] - alpha[2]]*f[-2*alpha[1] - alpha[2]]*v[Lambda[1]]^*
      - 1/2*f[-alpha[1]]*f[-3*alpha[1] - 2*alpha[2]]*v[Lambda[1]]^*
      - 1/2*f[-alpha[1] - alpha[2]]*f[-3*alpha[1] - alpha[2]]*v[Lambda[1]]^*
      + f[-2*alpha[1] - alpha[2]]^2*v[Lambda[1]]^*]
Got:
    [v[Lambda[1]]^*,
     f[-alpha[1]]*v[Lambda[1]]^*,
     f[-alpha[1] - alpha[2]]*v[Lambda[1]]^*,
     f[-alpha[1]]*f[-alpha[1] - alpha[2]]*v[Lambda[1]]^* + f[-2*alpha[1] - alpha[2]]*v[Lambda[1]]^*,
     f[-alpha[1]]^2*f[-alpha[1] - alpha[2]]*v[Lambda[1]]^* + f[-alpha[1]]*f[-2*alpha[1] - alpha[2]]*v[Lambda[1]]^* + 1/2*f[-3*alpha[1] - alpha[2]]*v[Lambda[1]]^*,
     f[-alpha[2]]*f[-3*alpha[1] - alpha[2]]*v[Lambda[1]]^* - 2*f[-alpha[1] - alpha[2]]*f[-2*alpha[1] - alpha[2]]*v[Lambda[1]]^* + f[-3*alpha[1] - 2*alpha[2]]*v[Lambda[1]]^*,
     f[-alpha[1]]*f[-alpha[2]]*f[-3*alpha[1] - alpha[2]]*v[Lambda[1]]^* - 2*f[-alpha[1]]*f[-alpha[1] - alpha[2]]*f[-2*alpha[1] - alpha[2]]*v[Lambda[1]]^* + f[-alpha[1]]*f[-3*alpha[1] - 2*alpha[2]]*v[Lambda[1]]^* + f[-alpha[1] - alpha[2]]*f[-3*alpha[1] - alpha[2]]*v[Lambda[1]]^* - 2*f[-2*alpha[1] - alpha[2]]^2*v[Lambda[1]]^*]
**********************************************************************
3 items had failures:
   2 of   7 in sage.algebras.lie_algebras.bgg_dual_module.BGGDualModule.degree_on_basis
   1 of   5 in sage.algebras.lie_algebras.bgg_dual_module.SimpleModule.lift
   2 of  14 in sage.algebras.lie_algebras.bgg_dual_module.SimpleModuleIndices.__contains__
    [265 tests, 5 failures, 26.58 s]
**********************************************************************
File "src/sage/algebras/lie_algebras/lie_algebra_element.pyx", line 967, in sage.algebras.lie_algebras.lie_algebra_element.UntwistedAffineLieAlgebraElement._repr_generic
Failed example:
    elt._repr_generic(str, str, lambda t: "T^{}".format(t), '.', '(x)')
Expected:
    '(E[alpha[3]] + E[alpha[2]] + E[alpha[1]] + h1 + h2 + h3 + E[-alpha[3]]
     + E[-alpha[2]] + E[-alpha[1]])(x)T^0 + (E[-alpha[1] - 2*alpha[2]
     - 2*alpha[3]])(x)T^1 + (E[alpha[1] + 2*alpha[2] + 2*alpha[3]])(x)T^-1 + c + d'
Got:
    '(E[alpha[1]] + E[alpha[3]] + E[alpha[2]] + h1 + h2 + h3 + E[-alpha[1]] + E[-alpha[3]] + E[-alpha[2]])(x)T^0 + (E[-alpha[1] - 2*alpha[2] - 2*alpha[3]])(x)T^1 + (E[alpha[1] + 2*alpha[2] + 2*alpha[3]])(x)T^-1 + c + d'
**********************************************************************
1 item had failures:
   1 of   4 in sage.algebras.lie_algebras.lie_algebra_element.UntwistedAffineLieAlgebraElement._repr_generic
    [463 tests, 1 failure, 1.23 s]
**********************************************************************
File "src/sage/algebras/lie_algebras/poincare_birkhoff_witt.py", line 545, in sage.algebras.lie_algebras.poincare_birkhoff_witt.PoincareBirkhoffWittBasis.casimir_element
Failed example:
    C = U.casimir_element(); C
Expected:
    1/4*PBW[alpha[2]]*PBW[-alpha[2]] + 1/12*PBW[alpha[1]]*PBW[-alpha[1]]
     + 1/12*PBW[alpha[1] + alpha[2]]*PBW[-alpha[1] - alpha[2]] + 1/12*PBW[2*alpha[1] + alpha[2]]*PBW[-2*alpha[1] - alpha[2]]
     + 1/4*PBW[3*alpha[1] + alpha[2]]*PBW[-3*alpha[1] - alpha[2]]
     + 1/4*PBW[3*alpha[1] + 2*alpha[2]]*PBW[-3*alpha[1] - 2*alpha[2]]
     + 1/12*PBW[alphacheck[1]]^2 + 1/4*PBW[alphacheck[1]]*PBW[alphacheck[2]]
     + 1/4*PBW[alphacheck[2]]^2 - 5/12*PBW[alphacheck[1]] - 3/4*PBW[alphacheck[2]]
Got:
    1/12*PBW[alpha[1]]*PBW[-alpha[1]] + 1/4*PBW[alpha[2]]*PBW[-alpha[2]] + 1/12*PBW[alpha[1] + alpha[2]]*PBW[-alpha[1] - alpha[2]] + 1/12*PBW[2*alpha[1] + alpha[2]]*PBW[-2*alpha[1] - alpha[2]] + 1/4*PBW[3*alpha[1] + alpha[2]]*PBW[-3*alpha[1] - alpha[2]] + 1/4*PBW[3*alpha[1] + 2*alpha[2]]*PBW[-3*alpha[1] - 2*alpha[2]] + 1/12*PBW[alphacheck[1]]^2 + 1/4*PBW[alphacheck[1]]*PBW[alphacheck[2]] + 1/4*PBW[alphacheck[2]]^2 - 5/12*PBW[alphacheck[1]] - 3/4*PBW[alphacheck[2]]
**********************************************************************
1 item had failures:
   1 of   8 in sage.algebras.lie_algebras.poincare_birkhoff_witt.PoincareBirkhoffWittBasis.casimir_element
    [176 tests, 1 failure, 5.25 s]
**********************************************************************
File "src/sage/algebras/lie_algebras/verma_module.py", line 1356, in sage.algebras.lie_algebras.verma_module.VermaModuleHomset.singular_vector
Failed example:
    v = H.singular_vector(); v
Expected:
    f[-alpha[2]]*f[-alpha[1]]^3*v[Lambda[1] - Lambda[3]]
     + 3*f[-alpha[1]]^2*f[-alpha[1] - alpha[2]]*v[Lambda[1] - Lambda[3]]
Got:
    f[-alpha[1]]^3*f[-alpha[2]]*v[Lambda[1] - Lambda[3]]
**********************************************************************
1 item had failures:
   1 of  47 in sage.algebras.lie_algebras.verma_module.VermaModuleHomset.singular_vector
    [422 tests, 1 failure, 3.91 s]
**********************************************************************
File "src/sage/algebras/lie_conformal_algebras/lie_conformal_algebra_element.py", line 170, in sage.algebras.lie_conformal_algebras.lie_conformal_algebra_element.LCAStructureCoefficientsElement._repr_
Failed example:
    R.2.T()+3*R.3
Expected:
    TB[alpha[1]] + 3*B[alpha[2] + alpha[3]]
Got:
    TB[alpha[2]] + 3*B[alpha[1] + alpha[2]]
**********************************************************************
1 item had failures:
   1 of   7 in sage.algebras.lie_conformal_algebras.lie_conformal_algebra_element.LCAStructureCoefficientsElement._repr_
    [31 tests, 1 failure, 0.39 s]
**********************************************************************
File "src/sage/algebras/lie_algebras/classical_lie_algebra.py", line 1999, in sage.algebras.lie_algebras.classical_lie_algebra.LieAlgebraChevalleyBasis.degree_on_basis
Failed example:
    [L.degree_on_basis(m) for m in L.basis().keys()]
Expected:
    [alpha[2], alpha[1], alpha[1] + alpha[2],
     2*alpha[1] + alpha[2], 3*alpha[1] + alpha[2],
     3*alpha[1] + 2*alpha[2],
     0, 0,
     -alpha[2], -alpha[1], -alpha[1] - alpha[2],
     -2*alpha[1] - alpha[2], -3*alpha[1] - alpha[2],
     -3*alpha[1] - 2*alpha[2]]
Got:
    [alpha[1],
     alpha[2],
     alpha[1] + alpha[2],
     2*alpha[1] + alpha[2],
     3*alpha[1] + alpha[2],
     3*alpha[1] + 2*alpha[2],
     0,
     0,
     -alpha[1],
     -alpha[2],
     -alpha[1] - alpha[2],
     -2*alpha[1] - alpha[2],
     -3*alpha[1] - alpha[2],
     -3*alpha[1] - 2*alpha[2]]
**********************************************************************
1 item had failures:
   1 of   3 in sage.algebras.lie_algebras.classical_lie_algebra.LieAlgebraChevalleyBasis.degree_on_basis
    [291 tests, 1 failure, 420.21 s]
**********************************************************************
File "src/sage/categories/regular_crystals.py", line 841, in sage.categories.regular_crystals.RegularCrystals.ElementMethods.dual_equivalence_class
Failed example:
    G.edges(sort=True,sort_vertices=True)
Expected:
    [([[1, 3, 4], [2, 5]], [[1, 3, 5], [2, 4]], 4),
     ([[1, 3, 4], [2, 5]], [[1, 2, 4], [3, 5]], 2),
     ([[1, 2, 5], [3, 4]], [[1, 3, 5], [2, 4]], 2),
     ([[1, 2, 5], [3, 4]], [[1, 3, 5], [2, 4]], 3),
     ([[1, 2, 3], [4, 5]], [[1, 2, 4], [3, 5]], 3),
     ([[1, 2, 3], [4, 5]], [[1, 2, 4], [3, 5]], 4)]
Got:
    [([[1, 3, 5], [2, 4]], [[1, 2, 5], [3, 4]], 2), ([[1, 3, 5], [2, 4]], [[1, 2, 5], [3, 4]], 3), ([[1, 3, 4], [2, 5]], [[1, 3, 5], [2, 4]], 4), ([[1, 3, 4], [2, 5]], [[1, 2, 4], [3, 5]], 2), ([[1, 2, 3], [4, 5]], [[1, 2, 4], [3, 5]], 3), ([[1, 2, 3], [4, 5]], [[1, 2, 4], [3, 5]], 4)]
**********************************************************************
1 item had failures:
   1 of   7 in sage.categories.regular_crystals.RegularCrystals.ElementMethods.dual_equivalence_class
    [127 tests, 1 failure, 0.24 s]
**********************************************************************
File "src/sage/combinat/posets/poset_examples.py", line 1491, in sage.combinat.posets.poset_examples.Posets.YoungsLatticePrincipalOrderIdeal
Failed example:
    P.cover_relations()
Expected:
    [[[], [1]],
     [[1], [1, 1]],
     [[1], [2]],
     [[1, 1], [2, 1]],
     [[2], [2, 1]],
     [[2, 1], [2, 2]]]
Got:
    [[[], [1]],
     [[1], [2]],
     [[1], [1, 1]],
     [[2], [2, 1]],
     [[1, 1], [2, 1]],
     [[2, 1], [2, 2]]]
**********************************************************************
1 item had failures:
   1 of   4 in sage.combinat.posets.poset_examples.Posets.YoungsLatticePrincipalOrderIdeal
    [181 tests, 1 failure, 3.24 s]
----------------------------------------------------------------------
sage -t --long --random-seed=0 src/sage/algebras/lie_algebras/bgg_dual_module.py  # 5 doctests failed
sage -t --long --random-seed=0 src/sage/algebras/lie_algebras/lie_algebra_element.pyx  # 1 doctest failed
sage -t --long --random-seed=0 src/sage/algebras/lie_algebras/poincare_birkhoff_witt.py  # 1 doctest failed
sage -t --long --random-seed=0 src/sage/algebras/lie_algebras/verma_module.py  # 1 doctest failed
sage -t --long --random-seed=0 src/sage/algebras/lie_conformal_algebras/lie_conformal_algebra_element.py  # 1 doctest failed
sage -t --long --random-seed=0 src/sage/algebras/lie_algebras/classical_lie_algebra.py  # 1 doctest failed
sage -t --long --random-seed=0 src/sage/categories/regular_crystals.py  # 1 doctest failed
sage -t --long --random-seed=0 src/sage/combinat/posets/poset_examples.py  # 1 doctest failed
----------------------------------------------------------------------

cyrilbouvier · 2024-09-10T11:28:28Z

@vbraun I really do not know what to do about these tests. Were they passing before the PR ?

For example the first one in your list check the return value of a method an_element() which seems to return a different value on 32 and 64 architectures. I did not write this test, I just modified it because the output was dependent on the internal representation of the Graph class (changed by this PR). But I know nothing about Lie algebras and I do not know how to fix it.

dcoudert · 2024-09-10T11:43:53Z

the doctest in src/sage/combinat/posets/poset_examples.py can be fixed with a sort. For the other ones, we can call for help on sage-level.

vbraun · 2024-09-10T20:38:45Z

Test were passing before this PR

If an_element returns different values you can just use the # 64 bit / # 32 bit markers, and then run subsequent test with a particular element.

Together with sorting output this would probably fix most

cyrilbouvier · 2024-09-11T07:31:12Z

the doctest in src/sage/combinat/posets/poset_examples.py can be fixed with a sort. For the other ones, we can call for help on sage-level.

done in 23494ba

For the tests in "src/sage/categories/regular_crystals.py", line 841, I think this is due to a bug in the class EdgesView of graphs: the tests asks for the vertices to be sorted G.edges(sort=True,sort_vertices=True)
But the first returned edge is not sorted

u, v, _ = ([[1, 3, 5], [2, 4]], [[1, 2, 5], [3, 4]], 2)
u < v  # should be True
False

Can someone confirm ?

Edit:
Something unclear is happening with sort_vertices: it can sort uncomparable items

sage: H = Graph([[1, 'a'],[ (1, 'a')]])
sage: H.edges(sort_vertices=True)
[(1, 'a', None)]

If sorting were really performed, an exception should have been raised.
I suspect that the sorting is done based on the indexes of the vertices in the internal representation of the graph.

I am looking into the others.

I disabled the two tests that depend on `an_element` method on 32-bit: one should not write tests depending on "random" element.

cyrilbouvier · 2024-09-11T08:29:48Z

Commit 8627957 should fix the remaining tests.

I used ... # 32-bit to disabled tests that I could not fix immediatly with a sort.
If it is not an accepted way of dealing with these problems and it prevent this PR to be merged, so be it, but I will not spend hours fixing badly written tests* in a part of SageMath I know nothing about and implementing math I do not understand.

(*) the failing tests relies an the fact that an element returned by the an_element method is always the same but the returned element depends in fact of the internal representation of object in a completly different part of SageMath.

For me the remaining problem is the failing doctest in "src/sage/categories/regular_crystals.py". But as I say earlier I suspect a bug in EdgesView. If it is confirm, I can open a ticket for this bug, but I will not try to fix it in this PR.

dcoudert · 2024-09-11T09:53:33Z

Edit: Something unclear is happening with sort_vertices: it can sort uncomparable items
sage: H = Graph([[1, 'a'],[ (1, 'a')]])
sage: H.edges(sort_vertices=True)
[(1, 'a', None)]
If sorting were really performed, an exception should have been raised. I suspect that the sorting is done based on the indexes of the vertices in the internal representation of the graph.

I am looking into the others.

in src/sage/graphs/base/sparse_graph.pyx, method _reorganize_edge, there is a try...except that tries to compare vertices and in case of failure silently pass.
so we can also get

sage: sage: H = Graph([[0, 1, 2, 'a'],[ ('a', 1)]])
....: sage: H.edges(sort_vertices=True)
[('a', 1, None)]

I don't know what to do here.

cyrilbouvier · 2024-09-16T13:38:38Z

I will close this PR and open a new one with less changes to avoid all the breaking doctests. Mainly I will remove commit bd63dd7 which removes 4 lines in StaticSparseBackend code that sort the vertices before creating the graph.
I still think this behaviour is a bug (sorting vertices of all immutable graphs should not be the default.), and I will open a ticket to document it but I will not try to fix it myself.

dcoudert · 2024-09-16T13:48:07Z

I agree, it's best to have a PR dedicated to the change in StaticSparseBackend and in method _reorganize_edge (which should raise an error when types are incomparable.

…rs but without the extra log complexity (2nd try)       This PR is based on the closed PR sagemath#38427 (by myself) that was breaking too much doctests. This PR improve the `init_short_digraph` function that is used to initialize `StaticSparseCGraph` (used for immutable `Graph` and `DiGraph`). Before, a boolean parameter `sort_neighbors` was used to specify if we wanted to sort the neighbors or not. It implied an extra `log` in the complexity (as `qsort` was called). With this PR, the neighbors are always sorted at no extra cost. It is done by appending to the neighbors list the vertices in the correct order so the call to `qsort` is not needed anymore. This PR partly reverts sagemath#38198 and mostly reverts sagemath#37662 Contrary to PR sagemath#38427, I did not include the patch that remove the sorting of vertices when StaticSparseGraph are initialized because it was breaking too many doctests. Instead I added a option to disabled said sorting and use it in my doctests to check that the new code works. ### 📝 Checklist  - [x] The title is concise and informative. - [x] The description explains in detail what this PR is about. - [ ] I have linked a relevant issue or discussion. - [x] I have created tests covering the changes. - [ ] I have updated the documentation and checked the documentation preview. ### ⌛ Dependencies    URL: sagemath#38664 Reported by: cyrilbouvier Reviewer(s): David Coudert

graphs: init_short_digraph always sort neighbors without extra log co…

b4556fa

…mplexity

github-actions bot added the s: needs review label Jul 26, 2024

cyrilbouvier marked this pull request as draft July 26, 2024 14:46

github-actions bot removed the s: needs review label Jul 26, 2024

mkoeppe added the c: graph theory label Aug 19, 2024

Remove sorting in StaticSparseBackend

bd63dd7

cyrilbouvier marked this pull request as ready for review August 21, 2024 07:40

github-actions bot added the s: needs review label Aug 21, 2024

Fix some tests whose output depend on the inner representation of sta…

71113cd

…tic graphs

Fix more tests whose output depend on the inner representation of sta…

07fc3e3

…tic graphs

dcoudert suggested changes Sep 4, 2024

View reviewed changes

cyrilbouvier added 3 commits September 4, 2024 18:54

Improve doctests syntax + respect linter + remove empty line

e22d5f1

Merge branch 'develop' into linear-sorted-init-short-digraph

7d085d9

Merge branch 'develop' into linear-sorted-init-short-digraph

d82009e

dcoudert approved these changes Sep 5, 2024

View reviewed changes

dcoudert added s: positive review and removed s: needs review labels Sep 5, 2024

vbraun added the s: needs work label Sep 8, 2024

Fix doctest by adding a sort

23494ba

github-actions bot added s: needs review and removed s: positive review s: needs work labels Sep 11, 2024

cyrilbouvier added 3 commits September 11, 2024 09:48

Maybe sorting will fix this doctests...

f742133

doctests in bgg_dual_module.py should passed now

6f4b168

I disabled the two tests that depend on `an_element` method on 32-bit: one should not write tests depending on "random" element.

doctests: more sort and disabling

8627957

cyrilbouvier closed this Sep 16, 2024

github-actions bot removed the s: needs review label Sep 16, 2024

cyrilbouvier mentioned this pull request Sep 16, 2024

graphs: make init_short_digraph always sort neighbors but without the extra log complexity (2nd try) #38664

Merged

5 tasks

cyrilbouvier mentioned this pull request Sep 16, 2024

graphs: StaticSparseBackend (i.e., immutable graphs) sorts the vertices by default #38665

Open

2 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

graphs: make init_short_digraph always sort neighbors but without the extra log complexity #38427

graphs: make init_short_digraph always sort neighbors but without the extra log complexity #38427

cyrilbouvier commented Jul 26, 2024

cyrilbouvier commented Jul 26, 2024

github-actions bot commented Jul 26, 2024 •

edited

Loading

cyrilbouvier commented Aug 19, 2024

cyrilbouvier commented Aug 21, 2024

cyrilbouvier commented Aug 22, 2024 •

edited

Loading

cyrilbouvier commented Aug 22, 2024

dcoudert left a comment •

edited

Loading

cyrilbouvier commented Sep 4, 2024

dcoudert left a comment

vbraun commented Sep 8, 2024

cyrilbouvier commented Sep 10, 2024

dcoudert commented Sep 10, 2024

vbraun commented Sep 10, 2024

cyrilbouvier commented Sep 11, 2024 •

edited

Loading

cyrilbouvier commented Sep 11, 2024 •

edited

Loading

dcoudert commented Sep 11, 2024 •

edited

Loading

cyrilbouvier commented Sep 16, 2024

dcoudert commented Sep 16, 2024

graphs: make init_short_digraph always sort neighbors but without the extra log complexity #38427

graphs: make init_short_digraph always sort neighbors but without the extra log complexity #38427

Conversation

cyrilbouvier commented Jul 26, 2024

📝 Checklist

⌛ Dependencies

cyrilbouvier commented Jul 26, 2024

github-actions bot commented Jul 26, 2024 • edited Loading

cyrilbouvier commented Aug 19, 2024

cyrilbouvier commented Aug 21, 2024

cyrilbouvier commented Aug 22, 2024 • edited Loading

cyrilbouvier commented Aug 22, 2024

dcoudert left a comment • edited Loading

Choose a reason for hiding this comment

cyrilbouvier commented Sep 4, 2024

dcoudert left a comment

Choose a reason for hiding this comment

vbraun commented Sep 8, 2024

cyrilbouvier commented Sep 10, 2024

dcoudert commented Sep 10, 2024

vbraun commented Sep 10, 2024

cyrilbouvier commented Sep 11, 2024 • edited Loading

cyrilbouvier commented Sep 11, 2024 • edited Loading

dcoudert commented Sep 11, 2024 • edited Loading

cyrilbouvier commented Sep 16, 2024

dcoudert commented Sep 16, 2024

github-actions bot commented Jul 26, 2024 •

edited

Loading

cyrilbouvier commented Aug 22, 2024 •

edited

Loading

dcoudert left a comment •

edited

Loading

cyrilbouvier commented Sep 11, 2024 •

edited

Loading

cyrilbouvier commented Sep 11, 2024 •

edited

Loading

dcoudert commented Sep 11, 2024 •

edited

Loading