subresultantSequence has a few problems, so
I replace subresultantSequence(p,q) with
subresultantVector(p,q).
Note that, degree(p)>degree(q) is always ture
in the changed part. And it used to use only last
degree(q) elements, now it uses first degree(q)
elementes.
--
You received this message
A grep shows that there are only 23 setrest! used in algebra
code (non data structure code). I suspect that none of them
will create a cyclic list. (I'll look into it). If so, we may make
List non-cyclic by default, and add a new domain that is cyclic.
> To put it differently: it seems that
oldk1331 wrote:
>
> I think the correct way is to have 2 domains for lists:
> one potentially with cycle, one guaranteed to not have cycle.
> Adding checks in setrest! should be enough to prevent
> the creation of cyclic lists.
Such check would have to travel to the end of added part
to be
oldk1331 wrote:
>
> BTW, the profile data is from SBCL profiler, right?
Yes.
> PS, we can add a "heuristicIntegrate", which tries varies
> human tricks to do integration, and keep "integrate" algorithmic.
Well 'heuristic' really mean "algortihm that we do not understand".
In case of
> Looking at IndexedList I see that 'copy' signals error on cyclic
> lists. It makes some sense as allowing 'copy' to run would exhaust
> memory and lead to crash. This probably motivates quite
> small value for 'cycleMax'. But if function just sits in
> infinite loop there not little problem:
oldk1331 wrote:
>
> > You mean we should have more special cases? In general
> > I am trying to have as little special cases as possible.
>
> > So possibly instead of seeking special
> > cases we should use modular methods.
>
> In this case, modular methods might be a good solution;
> but in
BTW, the profile data is from SBCL profiler, right?
PS, we can add a "heuristicIntegrate", which tries varies
human tricks to do integration, and keep "integrate" algorithmic.
--
You received this message because you are subscribed to the Google Groups
"FriCAS - computer algebra system" group.
> You mean we should have more special cases? In general
> I am trying to have as little special cases as possible.
> So possibly instead of seeking special
> cases we should use modular methods.
In this case, modular methods might be a good solution;
but in general case, integrate(c*f(x),x) =
oldk1331 wrote:
>
> > Yes. I am not sure how big is result of extendedEuclidean,
>
> It is not big. I computed it by resultantEuclidean in 50s,
> (semiResultantEuclidean2 in 30s).
You mean for the extendedEuclidean in
integrate((f^2 + x^2)/((x - a)* (x - b)* (x - c)* (x - d)), x)
> > -
Here's some doc and typo fixes I found recently.
--
You received this message because you are subscribed to the Google Groups
"FriCAS - computer algebra system" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to fricas-devel+unsubscr...@googlegroups.com.
= from URAGG checks for cyclic structure, while
= from List doesn't do this check. Is it for performance?
I think this is a bug.
diff --git a/src/algebra/list.spad b/src/algebra/list.spad
index 2c92001c..b228f395 100644
--- a/src/algebra/list.spad
+++ b/src/algebra/list.spad
@@ -116,7 +116,8 @@
On Sun, Mar 5, 2017 at 10:22 AM, Waldek Hebisch
wrote:
> As long as aggregate imposes order 'parts' should do:
> construct(parts(x)) shoud be x, so 'parts' should
> preserver whatever order aggregate has. Of course,
> if aggregate is unordered like Set, then 'parts'
>
oldk1331 wrote:
>
> OK, so there's a documentation bug in parts/members.
> And "entries" is available for aggregates which isn't
> a finiteAggregate. All of them don't guarantee order.
>
> What I want is a function that returns its elements to a
> list in order, so that I can convert/compare
>
"leaves" is not implemented for List/Stream, I add a
default one in RCAGG.
There's also a version for List that checks for loop
and a version for Stream that checks for infinity and loop.
--
You received this message because you are subscribed to the Google Groups
"FriCAS - computer algebra
OK, so there's a documentation bug in parts/members.
And "entries" is available for aggregates which isn't
a finiteAggregate. All of them don't guarantee order.
What I want is a function that returns its elements to a
list in order, so that I can convert/compare
IndexedVector/IndexedList etc. to
On 03/04/2017 09:51 PM, Waldek Hebisch wrote:
>> I think these 3 functions are the same.
> (1) -> s := multiset [1, 2, 3, 2, 3, 3, 3]
>
>(1) {1,2: 2,4: 3}
> Type: Multiset(PositiveInteger)
> (2) -> parts(s)
>
>(2) [1,2,2,3,3,3,3]
>
Martin R wrote:
>
> It would make sense to have one function that produces a list l such that
> "construct(l)@domain" produces the original thing.
For Collection 'parts' gives such list:
parts : % -> List S
++ parts(u) returns a list of the consecutive elements of u.
++
>
> I think these 3 functions are the same.
(1) -> s := multiset [1, 2, 3, 2, 3, 3, 3]
(1) {1,2: 2,4: 3}
Type: Multiset(PositiveInteger)
(2) -> parts(s)
(2) [1,2,2,3,3,3,3]
Type:
On 03/04/2017 12:35 PM, oldk1331 wrote:
> Ralf, could you update the GitHub mirror to include tag 1.3.1?
Sorry for the omission. It's pushed now.
Ralf
--
You received this message because you are subscribed to the Google Groups
"FriCAS - computer algebra system" group.
To unsubscribe from
It would make sense to have one function that produces a list l such that
"construct(l)@domain" produces the original thing.
Martin
--
You received this message because you are subscribed to the Google Groups
"FriCAS - computer algebra system" group.
To unsubscribe from this group and stop
Ralf, could you update the GitHub mirror to include tag 1.3.1?
--
You received this message because you are subscribed to the Google Groups
"FriCAS - computer algebra system" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to
I think these 3 functions are the same.
>From HOAGG, parts and members have the same
documentation: "returns a list of the consecutive
elements of u." "consecutive" doesn't imply "order".
>From IXAGG, entries "returns a list of all the entries
of aggregate u in no assumed order."
So all of
22 matches
Mail list logo