matplotlib-users Mailing List for matplotlib

On Mon, Jul 4, 2011 at 11:58 AM, Nico Schlömer <nic...@gm...>wrote:

> Also, set_{x,y,z}lim3d() doesn't seem to influence the plot at all. ?
>
> --Nico
>
>
>
Do you have an example demonstrating this?  Also, which version of
matplotlib are you using?  These functions should work as expected in
version  1.0.1.

Ben Root

On Mon, Jul 4, 2011 at 11:54 AM, Nico Schlömer <nic...@gm...>wrote:

> Hi all,
>
> this issue has come up before (see
>
> http://www.mail-archive.com/mat...@li.../msg14255.html
> )
> and now it was me who got bitten by it: As far as I can see, for
> mplot3d axes, it is not possible to set the aspect ratio of the axes
> to something other than 'auto' -- anyways, 'equal' does *not* make
> sure that the units of lenght on all axes are the same.
>
> For me, this functionality is important as I would like to highlight
> that certain angles are right angles which obviously isn't the case if
> the illustrations are distorted in that way.
>
> Cheers,
> Nico
>
>
Nico,

That is correct.  Due to the design of the mplot3d module, many functions
that are expected to work for 2D plots do not work for 3D plots.  I have
been working on bringing mplot3d to feature-pariety with the rest of mpl.
The upcoming release should address many of these issues.  However, the
set_aspect issue is a harder nut to crack correctly.

If you need a work-around, I have a git branch where I introduce the
matlab-esque pbaspect where one can specify the ratios of the x, y, and z
axes relative to each other.  Note that right now, at least one has to be
1.  It does introduce some other bugs, though, which is why it won't be in
the next release.

https://github.com/WeatherGod/matplotlib/tree/mplot3d%2Fpbaspect

Keep an eye on that branch for future updates, though.

I hope this helps,
Ben Root

Also, set_{x,y,z}lim3d() doesn't seem to influence the plot at all. ?

--Nico



On Mon, Jul 4, 2011 at 6:54 PM, Nico Schlömer <nic...@gm...> wrote:
> Hi all,
>
> this issue has come up before (see
> http://www.mail-archive.com/mat...@li.../msg14255.html)
> and now it was me who got bitten by it: As far as I can see, for
> mplot3d axes, it is not possible to set the aspect ratio of the axes
> to something other than 'auto' -- anyways, 'equal' does *not* make
> sure that the units of lenght on all axes are the same.
>
> For me, this functionality is important as I would like to highlight
> that certain angles are right angles which obviously isn't the case if
> the illustrations are distorted in that way.
>
> Cheers,
> Nico
>

Hi all,

this issue has come up before (see
http://www.mail-archive.com/mat...@li.../msg14255.html)
and now it was me who got bitten by it: As far as I can see, for
mplot3d axes, it is not possible to set the aspect ratio of the axes
to something other than 'auto' -- anyways, 'equal' does *not* make
sure that the units of lenght on all axes are the same.

For me, this functionality is important as I would like to highlight
that certain angles are right angles which obviously isn't the case if
the illustrations are distorted in that way.

Cheers,
Nico

On 1 July 2011 16:52, marz_cyclone <me...@me...> wrote:
> i'm trying to get the bounding box of a map plotted with basemap to place a
> colorbar. in this reduced example from the example directory of basemap, the
> colorbar is set to the hight of the axes of the plot.
>
...
>
> what i'm interested in is the bounding box of m, the real plotting area, so
> i can scale the colorbar axes to that height or width.

I don't have time to find the Basemap example data files and run your
example, but I normally do it like this (which works fine for me -
it's not clear what problem you're seeing?):

    # create new figure
    fig = plt.figure()
    ax = fig.add_axes(...)

    curr_map = Basemap(..., ax=ax, ...)

    im = curr_map.imshow(...)

    # setup colorbar axes and draw colorbar
    bbox = ax.get_position()
    l,b,w,h = bbox.bounds
    cax = fig.add_axes([l+w+0.05, b, 0.05, h],frameon=False)
    fig.colorbar(im, cax=cax)

Maybe using the pyplot functions to add the colorbar axes and plot the
colorbar is causing some side-effects you don't expect?

Cheers,
Scott