Re: No "square" plotting in 3d any more?!

[Ben Abbott]

On Dec 11, 2012, at 5:46 AM, address@hidden wrote:

> On 10/12/12 19:45:54, address@hidden wrote:
>> On 10/12/12 13:22:19, Jordi Gutiérrez Hermoso wrote:
>>> Also note that the viewpoint you pick for a 3d plot will induce a
>>> corresponding distortion in isometric projection.
>> 
> 
> Attached  an  image as I would expect it (this one was created from the former
> one by streching it in gimp, not really ideal). As the view  port  is  set  to
> view  along  the  diagonal  of the box, i.e. along the vector [1,1,1], I would
> expect that even in the projection all axes have the same length and therefore
> that the outline resembles a hexagon with equally long sides.
> 
> Many thanks for looking into this.
> Lynx

Please run this script ...

x = rand (100, 1);
y = rand (100, 1);
z = rand (100, 1);
figure (1)
clf ()
scatter3 (x, y, z, 10, 'b', "filled")

set (gca (), "plotboxaspectratio", [1 1 1])
filename = sprintf ("pbaspect=[%f,%f,%f].pdf", get (gca (), 
"plotboxaspectratio"));
filename(filename=="0") = [];
print ("-dpdfwrite", filename)

set (gca (), "plotboxaspectratio", [1 1 1.25])
filename = sprintf ("pbaspect=[%f,%f,%f].pdf", get (gca (), 
"plotboxaspectratio"));
filename(filename=="0") = [];
print ("-dpdfwrite", filename)

and email me the results.  I've attached the results I get.

pbaspect=[1.,1.,1.25].pdf
Description: Adobe PDF document

pbaspect=[1.,1.,1.].pdf
Description: Adobe PDF document

go_aspect.m
Description: Binary data