#15820: Implement sequences of bounded integers
-------------------------------------+-------------------------------------
Reporter: SimonKing | Owner:
Type: enhancement | Status: needs_review
Priority: major | Milestone: sage-6.3
Component: algebra | Resolution:
Keywords: sequence bounded | Merged in:
integer | Reviewers:
Authors: Simon King | Work issues:
Report Upstream: N/A | Commit:
Branch: | 5dc78c5f5f647bf95b98da916d7086fae539ffb8
u/SimonKing/ticket/15820 | Stopgaps:
Dependencies: |
-------------------------------------+-------------------------------------
Comment (by SimonKing):
Replying to [comment:52 ncohen]:
> - bounded integer sequence, biseq. What about "integer sequence" ?
Not good, because...
> All integers are "bounded" on a computer. Why add this "bounded"
everywhere ? biseq -> iseq
... the computer's integer bound is chosen when you buy the machine, but
not when you create the sequence.
The point is that each "bounded integer sequence" has a bound `B` that can
be ''chosen upon initialisation'' so that any item of the sequence is
smaller than `B`.
> - `# Bitsize of "data"` --> bitsize of the whole sequence
OK.
> - itembitsize --> necessarily a power of 2 ?
No. Here, it is really "bit" size. So, the number tells how many bits will
be reserved to store one item. Hence, the bound `B` above is a power of
two, namely `2^itembitsize`.
> - `#cdef inline void dealloc_biseq(biseq_t S)` : why there if commented
?
Because I forgot to remove it.
> - `list2biseq` (and others) -> `list_to_biseq` ?
Acceptable.
> - Comments on the function definitions --> Move it to the function's
doc, even
> if it is their only content ?
> - Could you write some doc, even if it is just one line, for cdef
functions ?
I wouldn't remove the comments from the pxd file, but certainly some
documentation should be added to the cdef functions in the pyx file. I
wonder if this should be by comments in the code or by a doc string. After
all, the doc string will not be visible (cdef functions won't appear in
the docs, as they can not imported).
> - I just figured out that you may have wanted to keep this list of
functions
> above with short descriptions as an 'index' of what the file does. If
so, it
> would be cool to do it in the html doc instead !
You mean: "This file provides the following boilerplate functions, that
you can cimport for usage in Cython code: ..."?
> - list2biseq --> I wondered why this function wouldn't return a new
biseq buil
> from a list instead of modifying an existing one.... Performance issue
? Is
> that critical in some cases ?
This is motivated by my usage in the `BoundedIntegerSequence` wrapper. As
usual, there is `__cinit__` which does memory allocations (as much as I
know, it is recommended to not do this kind of things in `__init__`).
Hence, when `__init__` is called (and I think filling data into the
allocated memory is the job of `__init__`, not of `__cinit__`), then the
integer sequence already is allocated, and just needs to be filled.
--
Ticket URL: <http://trac.sagemath.org/ticket/15820#comment:53>
Sage <http://www.sagemath.org>
Sage: Creating a Viable Open Source Alternative to Magma, Maple, Mathematica,
and MATLAB
--
You received this message because you are subscribed to the Google Groups
"sage-trac" 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 http://groups.google.com/group/sage-trac.
For more options, visit https://groups.google.com/d/optout.
