Thanks for the help. On Sunday, January 20, 2013, Andy Bauer wrote:
> FYI: this finally has made its way into ParaView such that it will be > available in the GUI for single process runs. > > Andy > > On Thu, Dec 20, 2012 at 8:08 PM, Sohail Shafii <sssha...@ucdavis.edu>wrote: > > Thanks. Please take your time. If there any other problems with this > reader then feel free to consult me. > > 1) typically I make a separate build directory for VTK or ParaView. The > test I ran looked for "../../../../vtkData" or something like that, so it > seems like the test executable has to be four directories deep. So I guess > there is a common parent to the data folders and paraview, not sure what > this parent would be or if there is a reference for this hierarchy. > > 2) I guess in this case I would have to experiment when I do make the > switch to the newer VTK > > > On Thu, Dec 20, 2012 at 9:45 AM, Andy Bauer <andy.ba...@kitware.com>wrote: > > Hi Sohail, > > Changing the number of fields from 23 to 16 fixed the warnings. I'll get > these changes into VTK and then change ParaView so that it can read in > WindBlade files with a single process. That may take a week or so though. > > As for the other questions: > > 1) If the data directories are called VTKData, VTKLargeData, and > ParaViewData and they share the same parent directory as VTK or ParaView, > when you build VTK or ParaView it will automatically pick up the path to > each of them. > > 2) For the library names, after modularization they've been changed and > relocated from the bin subdirectory to the lib subdirectory. Many of the > old libraries have been subdivided into better organized libraries. I don't > know any specific document that specifies what to do for this though. > > Andy > > > On Sun, Dec 16, 2012 at 9:39 PM, Sohail Shafii <sssha...@ucdavis.edu>wrote: > > I have some time now, so I addressed the issue. I don't the code has a > problem; the problem is related to the .wind file. > > Near the bottom, it describes how many fields exist in the comp.out files > ("DATA_VARIABLES"). For your test data, each file is about 64.5 MB. The > .wind says that each file should have 23 fields. For n scalar fields, 1 is > reserved for UVW; (n-1) for scalar data. So for this one, 1 for UWV while > 22 for scalar data. > > The UVW block is about 10 MB (160x80x70x(4 bytes)x(3 components)); each > scalar field is about 3.6 MB (160x80x70x(4 bytes)). So 10 MB + 22 scalar > fields*3.6 gives you about 90 MB. That's way too big compared to the > actual sizes of the comp.out files that you guys have for testing. * The > test works when you reduce the "DATA_VARIABLES" item from 23 to 16*(which > will give you close to 64.5 MB). > > I'm not sure why this .wind file is bad. Maybe older wind blade readers > just ignored errors like these. Anyway, the test seems to launch a window > and close right away. I didn't see any other errors after I fixed the wind > file that I am using. > > A couple of questions: > 1. How do you organize the VTK build directory with the testdata? The VTK > tests seem to expect the data to be four directories UP the hierarchy, so I > had to place my vtkData directory in my home folder. > 2. I tried the new versions of VTK with QT and it appears that the -l > commands that I put in the .pro file for linking (i.e. -lvtkIO) don't seem > to work...is there a document somewhere that helps in this regard? I > reverted to an older version of VTK for now but I want to eventually > upgrade. > > Sohail > > > On Sat, Nov 3, 2012 at 1:16 PM, Sohail Shafii <sssha...@ucdavis.edu>wrote: > > Also I'm removing Jimmy from this list since he may not be working with > wind files at this point. > > Sohail > > > On Sat, Nov > >
_______________________________________________ Powered by www.kitware.com Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html Please keep messages on-topic and check the ParaView Wiki at: http://paraview.org/Wiki/ParaView Follow this link to subscribe/unsubscribe: http://www.paraview.org/mailman/listinfo/paraview