Dear Andrzej,
First of all I feel really sorry that some simple examples from your
book are not supported anymore. To prevent such annoying thing in the
future, you could write all the doctests in a file and include them in
the dedicated repository "/tests/" of the Sage source code. Any changes
in Sage that break some of your doctest will immediately alert the
responsible programmer.
For concreteness, there was some recent changes in cartesian product. Mainly
http://trac.sagemath.org/ticket/18411
It cleans up a lot of things but at the same time is not fully
compatible anymore with old style parents (such as Combinations). In
order to solve it, the simplest would be to make Combinations a new
style parent and I created the following ticket for that purpose
http://trac.sagemath.org/ticket/19986
Best,
Vincent
On 30/01/16 09:07, Andrzej Giniewicz wrote:
Hello,
in earlier versions of sage this worked:
Omega = Combinations([1, 2, 3, 4, 5, 6], 4)
A1 = Combinations([1], 1)
A2 = Combinations([2, 3, 4, 5, 6], 3)
A = CartesianProduct(A1, A2)
A.cardinality() / Omega.cardinality()
which we used in book for high school students to illustrate basic
probability examples. Now this fails with assertion error (and no
explanation attached). Digging deeper gives:
File
"/usr/local/sage/sage-7.0/local/lib/python2.7/site-packages/sage/combinat/cartesian_product.py",
line 102, in CartesianProduct
return cartesian_product(iters)
File
"/usr/local/sage/sage-7.0/local/lib/python2.7/site-packages/sage/categories/cartesian_product.py",
line 174, in __call__
return super(CartesianProductFunctor, self).__call__(args, **kwds)
File
"/usr/local/sage/sage-7.0/local/lib/python2.7/site-packages/sage/categories/covariant_functorial_construction.py",
line 221, in __call__
assert(all( hasattr(arg, self._functor_name) for arg in args))
AssertionError
Simply making A1 and A2 list fails as well, because Combinations
iterates over lists, so list(A1) is list of lists.
TypeError: unhashable type: 'list'
Easiest way I found around this is:
A = cartesian_product([[tuple(a1) for a1 in A1], [tuple(a2) for a2 in A2]])
there is a LOT of syntax noise compared to simply:
A = CartesianProduct(A1, A2)
And the meaning is slightly different (in A there are tuples, in Omega
there are lists, so we are not comparing same objects).
Is there any simpler way to product output of Combinations? Limiting
syntax noise is very important for making mathematical packages
approachable to young students, and this change is definiately not
along the lines. Also, it is important that most use cases are with
same syntax, so if we have
A1 = [1,2,3]
A2 = [4,5,6]
A = cartesian_product([A1, A2])
it would be good if the syntax was same for Combinations and other objects.
Is the current behaviour designed or bug?
Regards,
Andrzej.
--
You received this message because you are subscribed to the Google Groups
"sage-support" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/sage-support.
For more options, visit https://groups.google.com/d/optout.
