doctest failure in sage 2.10.2.alpha1: const.tex

[aghitza]

Hi,

Computing the points on elliptic curves over finite fields goes through finding the structure of the corresponding abelian group, and the code picks random generators for this. Therefore the particular generators displayed, and the order in which the set of points is listed, vary from run to run.

This sometimes causes doctest failures on the examples in const.tex; I added comments containing the word "random" so that this doesn't happen.

Alex

Component: documentation

Issue created by migration from https://trac.sagemath.org/ticket/2238

[aghitza]

Attachment: const_fails.patch.gz

[williamstein]

comment:1

1. The right solution is to always SORT the output if possible (which it is), so there is nothing random involved.

2. And 2, my red flag went off when I saw list. Check out this bug!

sage: E = EllipticCurve(GF(5),[0, -1, 1, -10, -20]) 
sage: w = E.points(); w
[(0 : 1 : 0), (0 : 0 : 1), (1 : 4 : 1), (1 : 0 : 1), (0 : 4 : 1)]
sage: w[0] = 15
sage: E.points()
[15, (0 : 0 : 1), (1 : 4 : 1), (1 : 0 : 1), (0 : 4 : 1)]

I'll post a patch to fix both issues.

[williamstein]

this and doc-2238.patch should replace the above patch (apply this to sage repo; and doc to doc repo)

[williamstein]

Attachment: sage-2238.patch.gz

[williamstein]

comment:2

Attachment: doc-2238.patch.gz

[aghitza]

Attachment: sage-2238-new.patch.gz

replace sage-2238.patch

[aghitza]

comment:4

Looks great; unfortunately it breaks another doctest (trivial to fix, so I've done it and put in a new patch replacing sage-2238.patch). The doc-2238.patch looks fine.

[sagetrac-mabshoff]

comment:5

Merged sage-2238.patch and sage-2238-new.patch in Sage 2.10.2.alpah2

[JohnCremona]

comment:6

Thanks to Alex and William for fixing these things which were caused by me: as a relative novice to Python I am not aware as I should be of the critical differences between lists and tuples! And the need to sort things to be canonical.

I'm not sure what the correct milestone is for this, certainly not 2.10.2, and it's not in 2.10.3.rc0 either, but not for lack of positive reviews.

[sagetrac-mabshoff]

comment:7

Replying to @JohnCremona:

Thanks to Alex and William for fixing these things which were caused by me: as a relative novice to Python I am not aware as I should be of the critical differences between lists and tuples! And the need to sort things to be canonical.

I'm not sure what the correct milestone is for this, certainly not 2.10.2, and it's not in 2.10.3.rc0 either, but not for lack of positive reviews.

I check my 2.10.3.rc1 and:

• sage-2238-new.patch is "changeset: 8647:97976cb27ad3" in the sage repo (this is in the 2.10.2 release time frame)
• doc-2238.patch is "changeset: 429:3a30bbba6c29" in the doc repo (this is in the 2.10.2 release time frame)

So, I am sure both patches have been applied.

Cheers,

Michael