visualpython-users

[Visualpython-users] New Windows VPython

[Bruce S. <ba...@an...>]

At vpython.org is a new VPython for Windows that patches gaps in the 
Numeric module used in some VPython demo programs (and automatically 
imported by Visual). The update makes division of Numeric arrays work 
properly in the presence of "from __future__ import division". The patches 
to Numeric have been submitted to the keepers of the Numeric project.

Bruce Sherwood

[Visualpython-users] VPython for Python 2.2

[Bruce S. <ba...@an...>]

Today the final version of Python 2.2 was released. At vpython.org this is 
now the version supported for Windows, and we are working on bringing the 
Linux and Mac versions up to speed.

The new Windows version includes the new mouse and keyboard machinery (see 
the documentation), and zooming is now done with the middle button, making 
it possible to write programs which drag with the left button. (On the 
2-button mouse, the middle button is expressed by pressing both left and 
right buttons; on a 1-button mouse the middle button is expressed by 
holding down the CTRL key while holding the mouse button down.)

With the new version is a revised suite of demo programs that exploit the 
new mouse capabilities.

Note that with Python 2.2 you can put in your program the following 
statement:

	from __future__ import division

in which case 3/2 is 1.5 rather than truncated to 1. This is an important 
change for our physics students, who often trip over the truncation problem 
when doing scientific calculations. The intent of the Python developers is 
to make this behavior of division the default in Python 3.0, in the 
somewhat distant future.

Bruce Sherwood

[Visualpython-users] New VPython for Windows Python 2.2

[Bruce S. <ba...@an...>]

Another new experimental version of VPython for Windows Python 2.2 is 
available at vpython.org.

This package contains new versions of a number of demo programs, to exploit 
the new mouse machinery.

A significant change, soon to become standard, is that Visual no longer 
automatically imports the functions random, randint, and uniform from the 
whrandom module. This module has essentially been replaced by the random 
module. Rather than automatically import the newer random module, it seems 
prudent not to do this automatically. Some VPython users had already gotten 
caught by name conflicts due to this automatic import.

If you do have a program that uses random functions, you can insert this 
line to obtain the old behavior:

	from random import random, randint, uniform

Note however that randint is now officially "deprecated" in favor of a more 
Pythonish sequence-oriented randrange function.

It is important that Visual automatically imports the standard math module, 
so that trig and other functions are always available to all VPython 
programs. But the number of programs that require random functions is 
relatively small.

Bruce Sherwood

Re: [Visualpython-users] curve vs. cylinder

[Bruce S. <ba...@an...>]

I would expect cylinders to take longer to render because they really try 
to be real cylinders. Fat curves on the other hand are rendered as fast as 
possible and I think they have a square or diamond cross section.

Bruce Sherwood

--On Tuesday, December 18, 2001 12:30 PM -0500 art...@rs... wrote:

> My constructions tend to contain large numbers of lines, defined by 2
> points.
>
> Which gives me the option of using curve or cylinder.
>
> They seem to be quite distinct objects at the C code level,
> though I do seem to have a hard time distinguishing them
> visually.
>
> Does anyone have any insight into their differences, and which is likely
> to be more effiicent for rendering.
>
> Art

[Visualpython-users] more on mesh export to povray

[Arthur S. <aj...@ix...>]

From the Povray3.5(beta) help:

6.7.8  Triangle Texture Interpolation
This feature is utilized in a number of visualization approaches: triangles
with individual textures for each vertex, which are interpolated during
rendering.

So for Povray3.5 it should be possible to do an export from VPython that
preserves all features of VPython faces.

I am coming along with my effort.  Abandoned the concept of new keywords.
All faces exports will be a Povray mesh.
It will test for non-zero normals and create
smooth_triangles where they are found,
triangles otherwise.

I prefer to have explicitly declared textures
override colors, rather than combine with
them . So a declared texture will
trump any color info, and be included in
the mesh description.

If no declared texture:

Will test the objects colors.  If all vertex
colors are the same for all triangles will add the pigment at the mesh
level.

If not will test at triangle level. If all vertexes are the same color will
create
a texture (pigment) and a reference to it.
If vertexes have different colors, will add
a texture_list, and a reference to it.

Missing anything?

Art

[Visualpython-users] Improved mouse documentation

[Bruce S. <ba...@an...>]

At vpython.org the experimental Python 2.2 version of VPython for Windows 
has improved mouse documentation. Thanks to Ruth Chabay for suggestions.

Bruce Sherwood

[Visualpython-users] curve vs. cylinder

[<art...@rs...>]

My constructions tend to contain large numbers of lines, defined by 2 points.

Which gives me the option of using curve or cylinder.

They seem to be quite distinct objects at the C code level,
though I do seem to have a hard time distinguishing them
visually.

Does anyone have any insight into their differences, and which is likely to
be more effiicent for rendering.

Art

[Visualpython-users] Povray export of faces

[Arthur S. <aj...@ix...>]

FWIW I've done a workable faces_export def for the
povrayexport mod.

The only limitation I am aware of is in the handling of colors.

Faces allow per vertex coloring,  triangles in Povray do not - so that
limitation
is inherent.

My implementation also limits mesh objects to one color.  That can be
imrpoved
by creating texture objects for the individual triangles comprising the
mesh.

As implemented it adds 2 new export keywords"
   "face_meshes"  - to use the povray mesh primitive -  see povray docs for
advantages over
                            seperate triangles
   "face_normals" - whether to use triangle (without normals) or
smooth_triangels

So my exports def now has:

def export(display=None, filename=None, include_list=None, xy_ratio=4./3.,
custom_text='', shadowless=0, face_normals=1,face_mesh=0):

And, in iterating over the scene objects the amended code is:

        if function:
            if obj_name == "faces":
               object_code = function(obj,face_normals,face_mesh)
            else:
               object_code = function(obj)


Then:

def export_faces(a,normals,mesh):
    if normals:
       ttype="smooth_triangle"
    else:
       ttype="triangle"
    if mesh:
       object_template = """
mesh { %(triangles)s
       %(pigment)s }
         """
    triangle_template = """
%(ttype)s {
    <%(posx)f, %(posy)f, %(posz)f> %(n1)s,
    <%(pos2x)f, %(pos2y)f, %(pos2z)f> %(n2)s,
    <%(pos3x)f, %(pos3y)f, %(pos3z)f> %(n3)s
    %(pigment)s
}
   """
    triangles = []
    i=0
    if mesh :
       pigment = ""

    while i < len(a.pos):
       if normals:
          n1="<%f,%f,%f>" % (a.normal[i][0],a.normal[i][1],a.normal[i][2])
          n2="<%f,%f,%f>" %
(a.normal[i+1][0],a.normal[i+1][1],a.normal[i+1][2])
          n3="<%f,%f,%f>" %
(a.normal[i+2][0],a.normal[i+2][1],a.normal[i+2][2])
       else:
          n1=""
          n2=""
          n3=""

       if not mesh:
          pigment = "pigment { color rgb <%f, %f, %f> }" %tuple(a.color[i])

       triangles.append(triangle_template % {'ttype':ttype,
'posx':a.pos[i][0], 'posy':a.pos[i][1], 'posz':a.pos[i][2],
                                        'n1':n1,
                                        'pos2x':a.pos[i+1][0],
'pos2y':a.pos[i+1][1], 'pos2z':a.pos[i+1][2],
                                        'n2':n2,
                                        'pos3x':a.pos[i+2][0],
'pos3y':a.pos[i+2][1], 'pos3z':a.pos[i+2][2],
                                         'n3':n3,
                                         'pigment':pigment})
       i += 3

    triangles_in=""
    for j in triangles:
       triangles_in = triangles_in + j + '\n'

    if mesh:
       pigment = "pigment { color rgb <%f, %f, %f> }" %tuple(a.color[0])

       object_code = object_template %  { 'triangles':triangles_in,
                                         'pigment':pigment}
       object_code = add_texture(a, object_code)

    else:
       object_code =  triangles_in

    return object_code



Art

[Visualpython-users] VPython update for Python 2.2

[Bruce S. <ba...@an...>]

There is another update at vpython.org of VPython for Windows Python 2.2.

This fixes a mistake in handling zoom and spin which made these operations 
too sensitive to small mouse movements in large windows.

Bruce Sherwood

[Visualpython-users] Latest VPython for Python 2.2

[Bruce S. <ba...@an...>]

At vpython.org there is a VPython for Windows for the "candidate" version 
of Python 2.2, due out next week in final form.

This VPython package contains full documentation not only on the new mouse 
machinery but also on the new keyboard machinery implemented by Ari Heitner 
and David Andersen.

Bruce Sherwood

[Visualpython-users] new mouse facilities

[Bruce S. <ba...@an...>]

At vpython.org is a new Windows version for Python 2.2 which has a complete 
set of features for handling the mouse, including such things as 
scene.mouse.shift being true if the shift key is down at the time. See the 
documentation that comes with.

At python.org is an announcement that the final version of Python 2.2 will 
be released on Dec. 19. At that point we will change over completely to 
Python 2.2, and the new zoom convention will be the default (middle button 
to zoom rather than left button to zoom).

Bruce Sherwood

[Visualpython-users] VPython for Python 2.2

[Bruce S. <ba...@an...>]

At vpython.org is a new Windows VPython to go with the beta version of 
Python 2.2 (final release scheduled for Dec. 19).

This has new mouse machinery; see the on-line documentation. Though I've 
played with it quite a bit, feedback may lead to changes. It does maintain 
the old meaning of getclick(), so I think old programs should continue to 
work as before.

Thanks to work by Ari Heitner and David Andersen there is also machinery 
for reading the keyboard, but no documentation yet: the basic element is 
scene.kb.getkey() which gives you a character string suitable for printing.

Bruce Sherwood

Re: [Visualpython-users] strange error messages with F1

[Bruce S. <ba...@an...>]

I'm not sure I understand your final comment. Are you saying that later 
uses of F1 do get you the help, or rather that later uses of F1 don't do 
anything at all (including not giving you the error message)?

One oddity is the presence of the double backslashes in the URL. But maybe 
that's just an artifact of the error reporting, and the need in a character 
string to represent a backslash by two of them (since for example '\n' is 
end of line).

Bruce

--On Thursday, December 06, 2001 23:23 -0500 Joe Heafner 
<hea...@vn...> wrote:

> I'm running Python 2.1.1 and Visual 2001-12-04 under WinME on a Sony VAIO
> PCG-F630 notebook. When I press F1 from IDLE (VPython version) or select
> "Python Documentation" from the menu, I get the following in a debug
> window:
>
> Exception in Tkinter callback
> Traceback (most recent call last):
>   File "C:\PYTHON21\lib\lib-tk\Tkinter.py", line 1285, in __call__
>     return apply(self.func, args)
>   File "C:\Python21\Tools\idle_VPython\EditorWindow.py", line 327, in
>   python_docs webbrowser.open(self.help_url)
>   File "C:\PYTHON21\lib\webbrowser.py", line 43, in open
>     get().open(url, new, autoraise)
>   File "C:\PYTHON21\lib\webbrowser.py", line 221, in open
>     os.startfile(url)
> WindowsError: [Errno 2] The system cannot find the file specified:
> 'C:\\Python21\\Doc\\index.html'
>
>
> The help file is present and does indeed come up in Netscape, but I've
> never gotten these error messages before. Invoking F1 subsequent times
> does not generate any error messages. Any ideas?

[Visualpython-users] strange error messages with F1

[Joe H. <hea...@vn...>]

Hi.

I'm running Python 2.1.1 and Visual 2001-12-04 under WinME on a Sony VAIO PCG-F630 notebook. When I press F1 from IDLE (VPython version) or select "Python Documentation" from the menu, I get the following in a debug window:

Exception in Tkinter callback
Traceback (most recent call last):
  File "C:\PYTHON21\lib\lib-tk\Tkinter.py", line 1285, in __call__
    return apply(self.func, args)
  File "C:\Python21\Tools\idle_VPython\EditorWindow.py", line 327, in python_docs
    webbrowser.open(self.help_url)
  File "C:\PYTHON21\lib\webbrowser.py", line 43, in open
    get().open(url, new, autoraise)
  File "C:\PYTHON21\lib\webbrowser.py", line 221, in open
    os.startfile(url)
WindowsError: [Errno 2] The system cannot find the file specified: 'C:\\Python21\\Doc\\index.html'


The help file is present and does indeed come up in Netscape, but I've never gotten these error messages before. Invoking F1 subsequent times does not generate any error messages. Any ideas?

 

Cheers,
Joe

CVAC Home Page <http://users.vnet.net/heafnerj/cvac.html>
My Book <http://www.willbell.com/new/fundephcomp.htm>
My Home Page <http://users.vnet.net/heafnerj/>
Please -- no Microsoft attachments. They're a security risk.

[Visualpython-users] new povexport

[ruth c. <rc...@an...>]

Markus Gritsch has made a new version of povexport which automatically 
saves the generated code to a file.  Besides being more convenient, this 
produces significant speed improvements for large scenes.  The new version 
is available through the VPython website.  Thanks, Markus!

Re: [Visualpython-users] OS X version

[David A. <dm...@an...>]

I started on a Carbonized version, but then put it aside for other things. 
No idea when it will be out.

Re: [Visualpython-users] OS X version

[Bruce S. <ba...@an...>]

We would definitely like to have an OSX version, and some work is being 
done on this, but we have no estimate as to when it might be available. 
It's a fairly difficult problem.

Bruce Sherwood

--On Wednesday, December 05, 2001 13:23 -0600 Young-Jin Lee 
<yl...@ui...> wrote:

> I'd like to know where Mac OS X version of VPytnon stands.
> Will it come out? If so, when it will be?
> Thanks in advance.
>
> YJ

[Visualpython-users] OS X version

[Young-Jin L. <yl...@ui...>]

I'd like to know where Mac OS X version of VPytnon stands.
Will it come out? If so, when it will be?
Thanks in advance.

YJ

[Visualpython-users] physics applications

[Rob S. <sa...@ph...>]

I have collected some links to
VPython applications for Teaching Physics
at
http://www.physics.syr.edu/~salgado/software/vpython/
(see the bottom of that page)

If you know of any more, please let me know.


On that page, you can also find some of my attempts to use VPython
for visualizing electromagnetic fields and relativity.

rob
sa...@ph...

[Visualpython-users] www.vpython.org

[Bruce S. <ba...@an...>]

The VPython web site is now located at

	www.vpython.org (or simply vpython.org)

The old site (cil.andrew.cmu.edu/projects/visual) will eventually go away, 
so you should change your bookmarks and any VPython links you have on your 
own web pages. The old site was hosted at the Center for Innovation in 
Learning at Carnegie Mellon, which will close next summer.

Bruce Sherwood

Re: [Visualpython-users] can't bring up help files

[Bruce S. <ba...@an...>]

I don't understand some of this, but I did find that somehow the Windows 
VPython download contained the wrong version of 
Python21\Tools\idle_vpython. So I've rebuilt the VPython package.

--On Tuesday, December 04, 2001 21:05 -0500 Joe Heafner <hea...@vn...> 
wrote:

> Yesterday I installed Python 2.1.1 and the latest Visual (2001-11-18) on
> my WinME notebook. I accepted all default installation options (folders
> names, etc.) and the installation went smoothly. However, now F1 will not
> bring up the help files. Trying to access the help files from IDLE's menu
> doesn't work either. Pressing F1 repeatedly about 6 or 7 times causes
> IDLE to crash and generates an error in kernel.exe (kernel.dll?) and I
> have to restart the computer. Ctrl-alt-del doesn't even work. I've tried
> everything I can think of. F1 still works on all my desktop computers,
> and the EditorWindow.py file (?) is the same on all of the machines. I'm
> lost as to how to fix this.
>
> One thing that may be important. After installing Python 2.1.1, I noticed
> that the uninstall of Python 2.0 was not complete. The \Python20 folder
> tree had been left behind so I just dropped the whole tree into the
> recycle bin. I don't think this would have caused this problem though.

[Visualpython-users] ANN: PyGeo V.6

[Arthur S. <aj...@ix...>]

Congrats on the new VPython site.  Caught it in time
to correct the links at PyGeo before I went 
'final' with a newly done site, with newly
done PyGeo code.  

How 'bout giving a guy a link.  Math Forum found it.
And would *love* any feedback on how I'm using
 VPython.

And thanks to the VPython folks. PyGeo was
a shadow of its current self, pre-VPython . 
Had been using PyOpenGL.  But would have
never gotten to the performance, graphics
quality, Pov-Ray output, etc. left to 
my own resources.

Art

"Why don't you guys go do some coding!"
       Fredrik Lundh,  python-list, Dec 2, 2001
      (quote :approximate, spelling: exact)

And so I have.

BETTER CODE, REFRESHED SITE

http://home.netcom.com/~ajs

(check out the animation on the "links" page)
 
NEWS

PyGeo is a selected site at the Math Forum 

http://mathforum.org/library/topics/projective_g

From the PyGeo readme:

PyGeo is a 3d Dynamic Geometry toolset, written in Python,
with dependencies on Python's Numeric and VPython extensions.

It defines a set of geometric primitives in 3d space and allows 
for the construction of geometric models that can be interactively
manipulated, with defined geometric relationships remaining invariant.

It was created for, and is therefore particularly suitable for,
the visualization of concepts of Projective Geometry.

But it can used in more basic ways to create simple constructions
illustrating Euclidian principles.

And to create colorful designs from geometric ideas.

Fun (for all ages) as well as educational for both the
mathematically and *artistically* inclined.

It is also, hopefully:

 A take apart toy for folks trying to learn programming

 A dig at the "Old Dog,New Tricks" adage - having been created
 by a middle-ager without significant prior programming background.
 (It  took work, it took time)

 A work-in-process.
 
 
Art

[Visualpython-users] can't bring up help files

[Joe H. <hea...@vn...>]

Hi.

Yesterday I installed Python 2.1.1 and the latest Visual (2001-11-18) on my WinME notebook. I accepted all default installation options (folders names, etc.) and the installation went smoothly. However, now F1 will not bring up the help files. Trying to access the help files from IDLE's menu doesn't work either. Pressing F1 repeatedly about 6 or 7 times causes IDLE to crash and generates an error in kernel.exe (kernel.dll?) and I have to restart the computer. Ctrl-alt-del doesn't even work. I've tried everything I can think of. F1 still works on all my desktop computers, and the EditorWindow.py file (?) is the same on all of the machines. I'm lost as to how to fix this.

One thing that may be important. After installing Python 2.1.1, I noticed that the uninstall of Python 2.0 was not complete. The \Python20 folder tree had been left behind so I just dropped the whole tree into the recycle bin. I don't think this would have caused this problem though.

Any suggestions?


Cheers,
Joe

CVAC Home Page <http://users.vnet.net/heafnerj/cvac.html>
My Book <http://www.willbell.com/new/fundephcomp.htm>
My Home Page <http://users.vnet.net/heafnerj/>
Please -- no Microsoft attachments. They're a security risk.

[Visualpython-users] Newbie: VPython uses too much CPU time

[Record, J. N <jam...@lm...>]

I am running VPython with python 2.1.1 under RedHat 7.1. The X server uses
40-80% of the CPU time when the VPython display is active with no mouse
activity
or animation.

Is there a way to decrease the CPU usage? I am running a compute intensive
algorithm at the same time.

Jim Record
Lockheed Martin
303-977-0523
jam...@lm...

[Visualpython-users] [Q] How to create a derived class from a VPython object

[Young-Jin L. <yl...@ui...>]

I have been playing VPython and I think it is very easy to create a physics
learning tool.
I'd like to know if there is any way to create a derived class from a
VPython object.
I tried to create a derived class, Ball from a shpere, as follows, but it
didn't work.

from visual import *

class Ball(sphere):
    def __init__( self, pos, radius, color ):
        shpere.__init( self, pos, radius, color )
        self.name = 'Ball1'

    def setName( self, name ):
        self.name = name

    def getName( self ):
        return self.name


scene.width = scene.height = 500.
scene.x = scene.y = 0
scene.forward = ( -0.2, -0.5, -1 )
scene.autoscale = 0
scene.userzoom = 0
scene.range = ( 15, 15, 15 )

xmax = 10
dx = xmax / 5
grid = []
for x in arange( -xmax, xmax+dx, dx ):
    grid.append( curve( pos = [(x, 0, -xmax), (x, 0, xmax)],
                        color = ( .7, .7, .7 ) ) )

for z in arange( -xmax, xmax+dx, dx ):
    grid.append( curve( pos = [(-xmax, 0, z), (xmax, 0, z)],
                        color = (.7, .7, .7) ) )

ball = Ball(pos=(0,8,0), radius=0.5, color=color.red)
ball.velocity = vector(0,-1,0)
dt = 0.1
while ball.y > ball.radius:
    rate(100)
    ball.pos = ball.pos + ball.velocity*dt
    ball.velocity.y = ball.velocity.y - 9.8*dt
    wake = sphere( pos = ball.pos, radius = 0.1, color = color.blue )

When I ran the above script, I got the following error.
Traceback (innermost last)
  File
"c:\yjlee\worx\dissertation\python-related\pythonworx\conceptualphysics\free
fall.py", line 3, in ?
    class Ball(sphere):
TypeError: __init__() takes exactly 2 arguments (4 given)

I'm not sure if VPython object can be used a parent class because there is
no API document. (I couldn't find any information on the number of
parameters in the VPython object constructor.)

Thanks in advance.

Young-Jin Lee