[TRAC-3575] Strange conic behavoir

Thijs shared this problem 9 years ago
New

Hi,


I'm a new user and I haven't done much yet with Geogebra.

Recently I developed a "Clifford Torus" worksheet.

http://www.geogebratube.org/material/show/id/57647

Well isn't that nice?


In the proces of building I learned a lot of many beautiful ascepts of Geogebra.

It's really a great tool but I discovered a strange behavior of some conics.


Try for example:


c1=Conic[(0,-4),(0,4),(6,-8),(6.0000,8),(-2.0000,0)]

c2=Conic[(0,-4),(0,4),(6,-8),(6.0025,8),(-2.0004,0)]

c3=Conic[(0,-4),(0,4),(6,-8),(6.0025,8),(-2.0004166,0)]


Conic c1: Is correct.

Conic c2: Correctness of this conic depends on the position of axes within the graphic window.

Conic c3: Always incorrect.


These exceptional cases seem to indicate that something is caused by internal rounding errors.

Perhaps this behavior of conics is well known and accepted.

In that case, sorry for bothering.


If someone is interested:


For the rebel die-hards who refused to surrender:

See attached worksheet that shows incorrectness of some conics

caused by small displacement of certain points.

Don't waste too much time. These are exceptional cases.

https://ggbm.at/564357

Comments (2)

photo
1

Hi,


thanks for the report: https://jira.geogebra.org/b...


Cheers,

Zbynek

photo
1

Hi Kondr,


díky za přijetí mé žádosti.


I have no knowledge of Czech.

So I hope my answer makes sense.

Sorry for my poor English. :confused:


Cheers,

Thijs

© 2023 International GeoGebra Institute