Re: Another Simple-Plot Example

[Micha B. <kri...@us...>]

Tuesday, December 5, 2006, 22:32:33, Florian M=FCller wrote:

> The data field is calculated by some complex algebraic algorithms, but
> just for testing issues: Setting line 17 e.g. to

> newdata[i][j] =3D (double)(i*j);
> results in this output:

> http://students.informatik.uni-luebeck.de/~muellefl/unknown1.jpg

Unfortunately, I cannot see your row and column values. If I modify your =
source

  int columns =3D 40;
  int rows =3D 50;

I can see a similar picture (not twisted to the degree of your example).
Try to rotate it with your left mouse button and scale with your wheel.
I think, you will see something nicer than. Not actually the solution,
though.

> So that it seems, (at least for me) that it's not a data dependent prob=
lem.

It is in this example, because the z maxima climbs up to 2000 and the
aspect ratio of the poor thing turns from worse into ugly. You can
control it with Plot3D's scaling properties (also independent for all
three dimensions). The tic lines probably require some manual adaption
in such cases, but you have access to all of the 12 axes in order to
adjust them.
Another possibility is to pre-scale your data befor feeding them into
loadFromData.

> I'm using a Windows system with Qt 4.2.1, developing in Eclipse with CD=
T
> (i.e. compiling with MinGW).

Unfortunately I never tested it against this system. I hold a commercial
Qt license on Windows and don't bother there with g++ (probably I
should, for Qt4). BTW, the compiler option adheres to Visual C++, so you
can safely delete them.

I've used Qt4.2.2 + Visual Studio + qwtplot3d0.2.6 for this.

Micha =20
--=20