1)  There has been some discussion about this before.  Glyph creates a set
of geometry at each position in the field(s) that you pass to it.  If there
are a lot of positions and you request a "fancy" glyph, you get a lot of
geometry.  As you have noticed you can use a simpler glyph to lessen this
burden.  The other approach is to show glyphs at fewer points as you have
indicated.  There are three methods assuming the data have connections.
The first two assume regular connections (positions can be regular or
irregular).  Use one of them prior to Glyph or AutoGlyph
o  Reduce (interpolates the original grid to a lower resolution grid -- you
specify the reduction for each dimension, which can be different.for each,
the bounding box for the original data is invariant)
o  Slab (samples or subsets along a specific dimension)
o  Sample (samples spatially providing the number of points you specify,
but the results are scattered points -- no connections)

2) the mag function in Compute does this

3) by default glyph/autoglyph scales the length of vector glyphs by the
magnitude.  Using 2) you could also plot glyphs or other realizations of
the magnitude


Kate White <[EMAIL PROTECTED]>@opendx.watson.ibm.com on
06/22/2001 04:59:39 PM

Please respond to [email protected]

Sent by:  [EMAIL PROTECTED]


To:   <[email protected]>
cc:
Subject:  [opendx-users] glyphs



Hello all,

I am new to the opendx software, and I am having trouble with glyphs.  I
am trying to plot glyphs of a vector field and I have several questions:

1)  I keep running out of memory.  I changed the type of glyph, which
helped, but I would like to know how to plot only every tenth position or
so.

2)  I would like to use the compute module (or any module which works) to
take the length of each vector in the field and output a scalar field.

3)  If (2) can be accomplished, I would like to make the autoglyphs plot
vectors based on magnitude.

Thanks in advance,

Kate White
[EMAIL PROTECTED]



Reply via email to