2008/10/31 David E DeMarle <[EMAIL PROTECTED]>: > This is what I meant: > > 1) load the vtk file > ParaView now makes a pipeline that looks like this > > FileReader->Data > > And the data is like so (just a list of points in space): > x1,y1,z1: > x2,y2,z2: > ... > > 2) Apply the Calculator Filter to add a constant valued point > associated array to the data. > > Now the Pipeline looks like this: > FileReader->Data->Calculator->Data' > > And Data looks the same as it did, > While Data' is a set of locations in space, each with an associated value: > x1,y1,z1:someconstant > x1,y1,z1:someconstant > ... > > 3) save Data' to a file. Click on Calculator filter in pipelinebrowser > and select file->Save Data, and choose the VTK format. > > 4) compare the input and output files to see what the file should look > like to give each vertex file format for data with values > > Now that the data has some value, the glyph filter has something to > work with to let it decide how large to make the polygons that make up > each sphere. > > 5) load Data' into VTK, apply the glyph filter and turn scaling by > scalar value on. > > If the values meant something (like atomic number), instead of just > being a constant value, the glyph size would mean something. Without > that, the glyph filter just makes each sphere exactly 0.5 units wide > because it can not guess what you need it to look like.
OK, now it works! The newly-generated vtk files contains now an extra part POINT_DATA 5000 SCALARS Result double LOOKUP_TABLE default 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 where 1 is the value I gave to the calculator filter. I can then load the new vtk file, apply the gliph filter, set scale mode to "scalar", tick the edit box and set 1 as the value, then visualization is correct. > I am quite surprised that povray will make a guess actually. How does > it know what radius to choose that looks right? Are you trying to draw > stars from your list of coordinates, apples int a bin, marbles on a > table, bubbles in a pint of beer, or atoms in a molecule? Well, I cannot comment on Povray's inner workings. But to me the situation looks different: I am telling povray (or any other code at that regard) that in physical position (x1,y1,z1) I want a sphere with physical diameter 1 and in position (x2,y2,z2) another sphere always with diameter 1. If the distance between the two sphere [sqrt((x1-x2)^2 + (y1-y2)^2 + (z1-z2)^2))] is equal to one, the two spheres must touch, if larger they cannot and if smaller than 1 they overlap. How that is showed and rescaled on screen is not my business, but the fact that spheres overlap/do not overlap is fixed by their radius and their coordinates. I do not need (nor want) to provide povray with any extra info about what the sphere stands for. Thanks a lot for the very appreciated help. Kind Regards Lorenzo > > cheers, > Dave > > On Fri, Oct 31, 2008 at 1:08 PM, Lorenzo Isella > <[EMAIL PROTECTED]> wrote: >> Hello, >> I installed from source version 3.4.2. Now I can follow your procedure. >> I generated the new vtk file (saved as ascii), but it is definitely >> different from what I started from: >> >> # vtk DataFile Version 3.0 >> vtk output >> ASCII >> DATASET POLYDATA >> POINTS 250000 float >> 64.3441 20.798 26.378 64.3441 20.798 18.4467 66.0647 20.798 25.9853 >> 67.4446 20.798 24.8849 68.2103 20.798 23.2948 68.2103 20.798 21.5299 >> 67.4446 20.798 19.9398 66.0647 20.798 18.8394 65.5608 22.0147 25.9853 >> 66.5365 22.9904 24.8849 67.0779 23.5319 23.2948 67.0779 23.5319 21.5299 >> 66.5365 22.9904 19.9398 65.5608 22.0147 18.8394 64.3441 22.5187 25.9853 >> 64.3441 23.8985 24.8849 64.3441 24.6643 23.2948 64.3441 24.6643 21.5299 >> 64.3441 23.8985 19.9398 64.3441 22.5187 18.8394 63.1274 22.0147 25.9853 >> [...] >> >> For instance, now there are many more points. The generated text file >> is above 20Mb, so it is not a piece of cake to look for its new >> features. >> Cheers >> >> Lorenzo >> >> >> >> 2008/10/31 Lorenzo Isella <[EMAIL PROTECTED]>: >>> 2008/10/30 David E DeMarle <[EMAIL PROTECTED]>: >>>>> Alright, we'll get to the bottom of this. >>>>> This is a list of what I do (which is clearly NOT working). >>>>> (1) open the .VTK file I emailed you. >>>>> (2)click apply >>>>> (3) I cannot select the calculator filter right now in the GUI hence >>>> >>>> I can not replicate that problem. As soon as I hit apply the >>>> calculator filter is an option. I am on paraview 3.4. >>> >>> This unexpected. I am on Debian testing and I have paraview 3.2.2. >>> >>> >>>> >>>>> (4) I apply a gliph filter and set the radius to 0.1 >>>>> (5) I can now apply a calculator filter. I leave the defaults and >>>>> write 1 in the empty space before applying it >>>>> (6) nothing changes in the snapshot apart from the color >>>> >>>> It shouldn't. You have just associated some data with the polygonal >>>> spheres. You have not made them larger. >>>> >>>> The color changed because you now have some associated data, not just >>>> locations to work with, so paraview automatically colormaps the data. >>>> Your data is pretty boring (everything has the value 1.0), so the >>>> color is pretty lame, (everything is red). >>> >>> So this makes sense now. >>> >>> >>>>> (7) select again gliph filter, set sphere to radius 0.1 and click >>>>> apply (In the "scalar" field I now read "Result", so I thought >>>>> Paraview somehow had info about the previous filter I had applied), >>>> >>>> Yes, this second glyph filter now has something besides the x,y,z >>>> locations to work with. It has the same associated data that caused >>>> the color to change. >>>> >>>>> (8) Now everything is red and I have no idea of what I should do. >>>> >>>> At the bottom of Glyph2's Properties Tab of the Object Inspector, >>>> change scale mode to scalar. >>>> >>>> What you will now see is not what are aiming for, it actually shows a >>>> sphere on top of every vertex within the original sets of spheres. But >>>> change the calculator filters expression and see what happens. >>> >>> Here I am a bit flying blind. Changing the numerical value in the >>> filter (1, 5, 10) does not do anything at first sight. >>> However, if I save everything to a legacy ascii vtk file, I now get >>> something large (about 30Mb). >>> >>> >>>> You should figure out why the calculator couldn't be applied to the >>>> final_config.vtk file reader's output. >>> >>> This is really the bad news. Does anyone on this list use paraview >>> from Debian testing standard repositories? >>> Actually, I should add now the problem of visualizing correctly the >>> spheres appears even in the simple example with 50 particles only; >>> there I noticed that the particles (by using radius 0.5) are actually >>> slightly smaller than they should be (and povray also disagrees). >>> I am puzzled; I thought all this would take minutes. >>> Cheers >>> >>> Lorenzo >>> >>> >>>> >>>>> >>>>> What am I doing wrong? >>>>> Regards >>>>> >>>>> Lorenzo >>>>> >>>> >>>> >>>> >>>> -- >>>> David E DeMarle >>>> Kitware, Inc. >>>> R&D Engineer >>>> 28 Corporate Drive >>>> Clifton Park, NY 12065-8662 >>>> Phone: 518-371-3971 x109 >>>> >>> >>> >>> >>> -- >>> Free will does not consist in inverting the river flow, but in being >>> the fish that leaps upstream. >>> >> >> >> >> -- >> Free will does not consist in inverting the river flow, but in being >> the fish that leaps upstream. >> > > > > -- > David E DeMarle > Kitware, Inc. > R&D Engineer > 28 Corporate Drive > Clifton Park, NY 12065-8662 > Phone: 518-371-3971 x109 > -- Free will does not consist in inverting the river flow, but in being the fish that leaps upstream. _______________________________________________ ParaView mailing list ParaView@paraview.org http://www.paraview.org/mailman/listinfo/paraview