Purpose:
3d interactive visualization via OpenGL.
Description:
The gl tool does visualizes LAMMPS or ChemCell snapshots or data files and mesh files via OpenGL graphics calls. The interface to OpenGL is provided to Python by the open-source PyOpenGL toolkit. To do interactive visualization, consider using this tool in conjunction with the vcr tool.
The gl constructor takes a data object containing atom or mesh snapshots as an argument (dump, data, cdata, mdump).
The show() method displays a single image of the selected atoms or triangles of snapshot N and saves it as image.png. While an image is displayed, the view can be changed directly by using the mouse buttons in the OpenGl window to translate (left), rotate (middle), or zoom (right) the image. However, these mouse changes are not permanent, unless the corresponding trans(), rotate(), and zoom() methods are invoked. The all() method loops thru all selected snapshots, displaying each one in turn. The resulting image files are saved to image0000.png, image0001.png, etc. The prefix "image" can be changed via the file setting.
The bg(), size(), rotate(), trans(), zoom(), box(), label(), and nolabel() methods control various aspects of the images produced. Without the trans() and zoom() settings, the GL image should roughly fill the window and be centered.
Additional movie effects can be produced using the pan() and select() methods. The pan() method specifies an initial and final viewpoint that is applied to the images produced by the all() method. For intermediate images, the view parameters will be interpolated between their initial and final values. The pan() method can thus be used to rotate a single image or fly-by the simulation as it animates. The select() method performs additional atom selection for each image of the all() sequence. An image-dependent %g variable can be used in the select() string. The select() method can thus be used to slice thru the data set.
The acol(), arad(), bcol(), brad(), tcol(), and tfill() methods change attributes of the atoms, bonds, and triangles displayed. Each atom or bond returned from the data object has an integer "type" from 1 to N. The type is used to index into a list of RGB colors or radii for atoms and bond thickness. The adef(), bdef(), and tdef() methods setup default mappings of types to colors/radii. The other methods enable specific assignments to be made. The interpolation option (e.g. many types assigned to a few colors), enables a smooth rainbow of colors to be assigned to a range of types. Note that the dump tool allows any vector to be returned as an atom "type" via its atype setting. Thus displayed colors can be keyed to atom velocity or other properties.
Colors are specified with a string, e.g. "red". There are 140 pre-defined colors from this WWW page which can be examined by importing the "colors" variable from "vizinfo". New colors can be defined by assigning a nickname of your choice to an RGB triplet, as described below.
The visual quality of atom spheres can be set via the q() method. The axis() method toggles whether xyz axes are displayed in the OpenGL window. Red is the +x axis, green is +y, and blue is +z. The ortho() method toggles between an orthographic and perspective display of the snapshot. The clip() method can be used to narrow the amount of data that is visualized. The reload() method is needed if you change the selection attributes of the underlying data assigned to the gl tool, such as via the dump tool's methods. These changes will not be visible in the OpenGL window until the data is reloaded.
Usage:
g = gl(d) create OpenGL display for data in d g = gl(d,N) window of size NxN (def = 512x512) g = gl(d,N,M) window of size NxM
d = atom snapshot object (dump, data)
g.bg("black") set background color (def = "black")
g.rotate(60,135) view from z theta and azimuthal phi (def = 60,30)
g.shift(x,y) translate by x,y pixels in view window (def = 0,0)
g.zoom(0.5) scale image by factor (def = 1)
g.box(0/1/2) 0/1/2 = none/variable/fixed box
g.box(0/1/2,"green") set box color
g.box(0/1/2,"red",4) set box edge thickness
g.file = "image" file prefix for created images (def = "image")
g.show(N) show image of snapshot at timestep N
g.all() make images of all selected snapshots g.all(P) images of all, start file label at P g.all(N,M,P) make M images of snapshot N, start label at P
g.pan(60,135,1.0,40,135,1.5) pan during all() operation g.pan() no pan during all() (default)
args = z theta, azimuthal phi, zoom factor at beginning and end values at each step are interpolated between beginning and end values
g.select = "$x > %g*3.0" string to pass to d.aselect.test() during all() g.select = "" no extra aselect (default)
%g varies from 0.0 to 1.0 from beginning to end of all()
g.acol(2,"green") set atom colors by atom type (1-N) g.acol([2,4],["red","blue"]) 1st arg = one type or list of types g.acol(0,"blue") 2nd arg = one color or list of colors g.acol(range(20),["red","blue"]) if list lengths unequal, interpolate g.acol(range(10),"loop") assign colors in loop, randomly ordered
if 1st arg is 0, set all types to 2nd arg if list of types has a 0 (e.g. range(10)), +1 is added to each value interpolate means colors blend smoothly from one value to the next
g.arad([1,2],[0.5,0.3]) set atom radii, same rules as acol()
g.bcol() set bond color, same args as acol() g.brad() set bond thickness, same args as arad()
g.tcol() set triangle color, same args as acol() g.tfill() set triangle fill, 0 fill, 1 line, 2 both
g.lcol() set line color, same args as acol() g.lrad() set line thickness, same args as arad()
g.adef() set atom/bond/tri/line properties to default g.bdef() default = "loop" for colors, 0.45 for radii g.tdef() default = 0.25 for bond/line thickness g.ldef() default = 0 fill
by default 100 types are assigned if atom/bond/tri/line has type > # defined properties, is an error
from vizinfo import colors access color list print colors list defined color names and RGB values colors["nickname"] = [R,G,B] set new RGB values from 0 to 255
140 pre-defined colors: red, green, blue, purple, yellow, black, white, etc
Settings specific to gl tool:
g.q(10) set quality of image (def = 5)
g.axis(0/1) turn xyz axes off/on
g.ortho(0/1) perspective (0) vs orthographic (1) view
g.clip('xlo',0.25) clip in xyz from lo/hi at box fraction (0-1)
g.reload() force all data to be reloaded
g.cache = 0/1 turn off/on GL cache lists (def = on)
theta,phi,x,y,scale,up = g.gview() grab all current view parameters
g.sview(theta,phi,x,y,scale,up) set all view parameters
data reload is necessary if dump selection is used to change the data
cache lists usually improve graphics performance
gview returns values to use in other commands:
theta,phi are args to rotate()
x,y are args to shift()
scale is arg to zoom()
up is a 3-vector arg to sview()
Related tools:
dump, rasmol, raster, svg, vcr, vmd
Prerequisites:
Python PyOpenGL and PIL packages.