Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
Drexel University or The Math Forum.
|
|
|
|
Re: Is there a way to visualize 4D data?
Posted:
Jun 26, 2006 4:17 PM
|
|
mixi.mo@virgin.net wrote:
> Lance wrote:
> > Perspective drawing drawing tries to make 3 dimensions fit onto two by
> > projecting then onto a two dimesnional surface.
> >
> > The same idea can be used but instead of projecting the imge you would
> > need to project the shadow - the three dimensional shadow of the four
> > dimensional object.
> >
> > Having said that I have no idea how you could actually do this with
> > your data.
> >
> > Lance
> >
> > gino wrote:
> > > I tried to do cross section, and slicing, on my 4D data, it is really
> > > awkward... My 4D data is actually sampled from z=f(x, y, u, v) in Matlab.
> > > Are there any tools existing in science and engineering field,
> > > that can help me visualize high dimensional data(using Matlab or other
> > > languages)?
> > >
> > > Thanks a lot!
>
> I found a similar discrepancy as Lance has. I always used
> point, line and plane (1st, 2nd, 3rd where the 4th would be
> = to geometries 3rd dimension) I can only pin it down as falling
> somewheres between philosophy and cognitive sciences.
>
> Using colour might be useful, but I think you might have to
> explain that the idea you present works maybe something
> like a stereogram? giving an extra dimension that wouldn't
> normally be there. M.
This isn't 4D by any means and a rather off-topic reply, but your
comment about stereograms reminded me of a fun thing you can try. Put a
set of identical 3D objects ( I used chess pawns last time) in a line
on a plain surface. Adjust the spacings ever so slightly and then
deconverge your visual focus as if viewing a single-image
autostereogram. You'll get a 3D effect. Now try moving a few of the
objects a tiny bit closer and further away in Z. Also try strong
directional lighting. Not *very* interesting but amusing.
|
|
|
|
