Opened 8 years ago
Closed 8 years ago
#5879 closed enhancement (fixed)
[with patch, positive review] Added crystal of letters for type E
Reported by: | aschilling | Owned by: | aschilling |
---|---|---|---|
Priority: | major | Milestone: | sage-4.0 |
Component: | combinatorics | Keywords: | combinat, crystals |
Cc: | sage-combinat-commits | Merged in: | |
Authors: | Reviewers: | ||
Report Upstream: | Work issues: | ||
Branch: | Commit: | ||
Dependencies: | Stopgaps: |
Description (last modified by )
This patch adds crystal of letters for type E corresponding to the highest weight crystal B(\Lambda_1) and its dual B(\Lambda_6) (in the sage labeling convention of the Dynkin nodes).
This patch depends on ticket #5120.
Attachments (1)
Change History (15)
comment:1 Changed 8 years ago by
- Description modified (diff)
comment:2 follow-up: ↓ 4 Changed 8 years ago by
comment:3 Changed 8 years ago by
- Summary changed from [with patch, needs review] Added crystal of letters for type E to [with patch, favorable review] Added crystal of letters for type E
After some testing I am confident that the crystal is correct.
Adding this one crystal (with its dual) actually gives access to all E6 crystals (modulo computational complexity) since they can be obtained as subcrystals of tensor products of these two. For example, the second fundamental weight corresponds to a representation of degree 351 whose crystal can be obtained thus:
sage: C = CrystalOfLetters("E6") sage: T = TensorProductOfCrystals(C,C,generators=[[C[0],C[0]]]) sage: T.cardinality() 351
The fact that this process (using both crystals) is complete to give all crystals is confirmed here:
http://groups.google.com/group/sage-combinat-devel/msg/8ef5b6b5b529e51b?hl=en
This crystal is also interesting because the action of the E6 Weyl group on this crystal of degree 27 must be equivalent to the action of E6 on the 27 lines on a cubic surface.
comment:4 in reply to: ↑ 2 Changed 8 years ago by
What is the logic of the representations of the elements? That is, why [-5,2,6]?
The nodes of this crystal are labeled by their weights. [-5,2,6] indicates that a 5-arrow is coming in, and a 2-arrow and 6-arrow is leaving the vertex.
I now also added a repr method that allows for concise printing of the nodes by +abcdefghijklmnopqrstuvwxyz and -ABCDEFGHIJKLMNOPQRSTUVWXYZ in the dual crystal.
comment:5 Changed 8 years ago by
- Summary changed from [with patch, favorable review] Added crystal of letters for type E to [with patch, positive review] Added crystal of letters for type E
The revision of the patch implementing concise representation is important since for tensor products of crystals the names of elements would be very cumbersome. The crystal at hand has highest weight the first fundamental weight, lambda1 in the Bourbaki notation. The dual crystal has highest weight lambda6:
O 2 | | O---O---O---O---O 1 3 4 5 6 E6
These both have degree 27. The adjoint representation has highest weight lambda2 has degree 78 and can be constructed as follows:
sage: C = CrystalOfLetters(['E',6], element_print_style = 'compact') sage: D = CrystalOfLetters(['E',6], element_print_style = 'compact', dual=True) sage: hwv=TensorProductOfCrystals(C,D).highest_weight_vectors(); hwv [[+, -], [j, -], [x, -]] sage: T = TensorProductOfCrystals(C,D,generators=[hwv[1]]) sage: T.cardinality() 78 sage: T.latex_file("/home/bump/tmp/e6-78.tex")
The last step assumes you have dot2tex installed. (Why isn't this bundled with SAGE?) Here is the resulting crystal of the adjoint representation:
comment:6 Changed 8 years ago by
Mhh, what is the credit situation here? "Brant Jones" was added to the list of authors, so is this shared credit for Anne and Brant for authorship?
Cheers,
Michael
comment:7 follow-up: ↓ 9 Changed 8 years ago by
Yes, shared authorship please.
Anne
comment:8 Changed 8 years ago by
- Summary changed from [with patch, positive review] Added crystal of letters for type E to [with patch, needs work] Added crystal of letters for type E
This breaks pickling for me (and #5120 by itself did not):
** failed: _class__sage_combinat_crystals_letters_ClassicalCrystalOfLetters__.sobj ** failed: _class__sage_combinat_crystals_tensor_product_CrystalOfTableauxElement__.sobj ** failed: _class__sage_combinat_crystals_tensor_product_CrystalOfTableaux__.sobj ** failed: _class__sage_combinat_crystals_tensor_product_FullTensorProductOfClassicalCrystals__.sobj ** failed: _class__sage_combinat_crystals_tensor_product_TensorProductOfClassicalCrystalsWithGenerators__.sobj doctest:1172: DeprecationWarning: RQDF is deprecated; use RealField(212) instead. Failed: _class__sage_combinat_crystals_letters_ClassicalCrystalOfLetters__.sobj _class__sage_combinat_crystals_tensor_product_CrystalOfTableauxElement__.sobj _class__sage_combinat_crystals_tensor_product_CrystalOfTableaux__.sobj _class__sage_combinat_crystals_tensor_product_FullTensorProductOfClassicalCrystals__.sobj _class__sage_combinat_crystals_tensor_product_TensorProductOfClassicalCrystalsWithGenerators__.sobj
Also:
from sage.quadratic_forms.extras import sgn
is going away since it is basically cmp(a,b).
Cheers,
Michael
comment:9 in reply to: ↑ 7 Changed 8 years ago by
Replying to aschilling:
Yes, shared authorship please.
Anne
Cool. Please specify shared authorship when you post the patch via the comment section during the file upload. That makes it just easier for me to keep track of credit.
Cheers,
Michael
comment:10 Changed 8 years ago by
I just talked to Nicolas about the pickling problem; this is a shortcoming of the current unique representation patch and he will try to find a solution to the problem in patch 5120.
Changed 8 years ago by
comment:11 Changed 8 years ago by
The revised patch fixes the sgn versus comparison issue. The pickling problem will hopefully be fixed by modifications in unique representations!
comment:12 Changed 8 years ago by
Nicolas is about to submit a new version of patch #5120 which should fix the pickling problem Michael mentioned above. We checked this by hand, but Michael, could you run your automatic pickling tests again?
comment:13 Changed 8 years ago by
- Summary changed from [with patch, needs work] Added crystal of letters for type E to [with patch, positive review] Added crystal of letters for type E
Pickling seems fine and #5120 got a positive review. So I reinstate the positive review.
comment:14 Changed 8 years ago by
- Milestone changed from sage-4.0.1 to sage-4.0
- Resolution set to fixed
- Status changed from new to closed
Merged in Sage 4.0.rc0.
Cheers,
Michael
For what it's worth, here's a picture of this crystal:
http://sporadic.stanford.edu/bump/xtal/e6.pdf
What is the logic of the representations of the elements? That is, why [-5,2,6]?