On Mon, Aug 24, 2009 at 11:54:20AM -0400, Tom Foutz wrote:
> I have a number of lines which are all part of a single segment.� When I
> plot them all individually, it takes a long time, much longer than
> plotting more complicated meshes.� Is there a way I can display the lines
> more efficiently?� Is there a way I can make these one object in mlab, so
> when I use the toolbar I don't get 30 different LineSource objects?
First of all, here is an improved version of you demo script, simply
coding thing a bit cleaner with Mayavi/numpy:
from enthought.mayavi import mlab
import numpy as np
mlab.clf()
# Number of lines
n_lines = 200
# Number of points per line
n_points = 100
# Create Example Coordinates
xyz=[]
for ii in xrange(n_lines):
xyz.append(
np.cumsum(np.random.random((n_points, 3)), axis=0)
)
fig = mlab.gcf()
fig.scene.disable_render = True
for this_xyz in xyz:
x, y, z = this_xyz
mlab.plot3d(x,
y,
z, tube_sides=7, tube_radius=0.1)
fig.scene.disable_render = False
Second, what you want is a graph: an object with a set of vertices
connected in an arbitrary way. There are two examples of graphs in the
Mayavi examples:
http://code.enthought.com/projects/mayavi/docs/development/html/mayavi/auto/example_protein.html#example-protein
http://code.enthought.com/projects/mayavi/docs/development/html/mayavi/auto/example_flight_graph.html#example-flight-graph
The idea is that you specify the points and then their connections, as in
index number of the flatten point array. This is a bit tedious, because
it is easy to get it wrong, but of course, it is more powerful than the
simple mlab.plot3d (that simply builds the connection array for you).
Here is an example on your demo:
from enthought.mayavi import mlab
import numpy as np
mlab.clf()
# Number of lines
n_lines = 200
# Number of points per line
n_points = 100
# Create Example Coordinates
xyz=[]
for ii in xrange(n_lines):
xyz.append(
np.cumsum(np.random.random((n_points, 3)), axis=0)
)
xyz = np.array(xyz)
x, y, z = xyz.T
pts = mlab.pipeline.scalar_scatter(x.T, y.T, z.T)
connections = []
for i in range(n_lines):
connections.append(np.c_[np.arange(i*n_points, (i+1)*n_points-1),
np.arange(i*n_points+1, (i+1)*n_points)])
pts.mlab_source.dataset.lines = np.vstack(connections)
# First option: tubes
if hasattr(mlab.pipeline, 'stripper'):
lines = mlab.pipeline.surface(
mlab.pipeline.tube(
mlab.pipeline.stripper(
pts,
),
tube_sides=7, tube_radius=0.1,
),
)
else:
lines = mlab.pipeline.surface(
mlab.pipeline.tube(
pts,
),
tube_sides=7, tube_radius=0.1,
),
)
> (Also, as you can see in the demo below, each segment has a little space
> between them.� Is there any way I can connect them better so there is no
> space between them? I was thinking about creating a sphere at each vertex
> to fill the gap.)
The right way to solve this is to use a stripper filter. It was added on
August 3rd to Mayavi, and is used automatically by points3d, but is not
in the released version. I added a check above to use it when available.
HTH,
Ga�l
------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
MayaVi-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mayavi-users
