non-unique kernels

[Ralf Hemmecke]

As you can see below, I must have been able to create an algebraic
number that looks like zero, but comes with non-equal kernels.

That shouldn't happen, right?

It's probably a problem with the cache of kernels, but why that is I do
not quite understand. Where should I take a closer look?

Ralf

%%% (552) -> hhh

+--+ +--+
- 325 \|33 + 325 \|33
(552) -----------------------
1179648
Type:
AlgebraicNumber
%%% (553) -> numer hhh

+--+ +--+
(553) - 325 \|33 + 325 \|33
Type:
SparseMultivariatePolynomial(Integer,Kernel(AlgebraicNumber))
%%% (554) -> kernels hhh

+--+ +--+
(554) [\|33 , \|33 ]
Type:
List(Kernel(AlgebraicNumber))
%%% (555) -> l := kernels hhh

+--+ +--+
(555) [\|33 , \|33 ]
Type:
List(Kernel(AlgebraicNumber))
%%% (556) -> l.1

+--+
(556) \|33
Type:
Kernel(AlgebraicNumber)
%%% (557) -> l.1=l.2

+--+ +--+
(557) \|33 = \|33
Type:
Equation(Kernel(AlgebraicNumber))
%%% (558) -> (l.1=l.2)@Boolean

(558) false
Type:
Boolean

[Qian Yun]

Did you ")compile" something (e.g. EXPR, AN) in this session?

- Qian

On 10/14/24 1:59 PM, 'Ralf Hemmecke' via FriCAS - computer algebra

[Ralf Hemmecke]

Thank you, Qian,

indeed, I compiled in this session. Without compiling, everything seems
to be fine.

Ralf

On 10/14/24 09:04, Qian Yun wrote:
> Did you ")compile" something (e.g. EXPR, AN) in this session?
>
> - Qian
>

> On 10/14/24 1:59 PM, 'Ralf Hemmecke' via FriCAS - computr algebra

[Waldek Hebisch]

On Mon, Oct 14, 2024 at 09:09:30AM +0200, 'Ralf Hemmecke' via FriCAS - computer algebra system wrote:
> Thank you, Qian,
>
> indeed, I compiled in this session. Without compiling, everything seems to
> be fine.
>
> Ralf
>
> On 10/14/24 09:04, Qian Yun wrote:
> > Did you ")compile" something (e.g. EXPR, AN) in this session?
> >
> > - Qian
> >
> > On 10/14/24 1:59 PM, 'Ralf Hemmecke' via FriCAS - computr algebra system
> > wrote:
> > > As you can see below, I must have been able to create an algebraic
> > > number that looks like zero, but comes with non-equal kernels.
> > >
> > > That shouldn't happen, right?
> > >
> > > It's probably a problem with the cache of kernels, but why that is I
> > > do not quite understand. Where should I take a closer look?
> > >
> > > Ralf
> > >
> > > %%% (552) -> hhh
> > >
> > >                   +--+        +--+
> > >            - 325 \|33  + 325 \|33
> > >     (552)  -----------------------
> > >                    1179648
> > >                                                          Type:
> > > AlgebraicNumber

Actual reason for this seem to be flushing of kernel cache
(this is done by compilation but also by ')clear all').
I added code which is supposed to handle such situations
and it mostly works. More precisely, when the cache is flushed
kerensl in the cache are specially marked by setting its position
to 0. When such kernels are later used in operations they should
be re-inserted in the cache. But apparently in some cases it does
not work...

--
Waldek Hebisch