Hi Carrie (cc. to Amos), Thanks a lot for reporting this - turned out to be a bug in reading the particle type from ascii files.
Now fixed (use "svn up" to get the latest). r5242 Daniel > On 18 Jun 2015, at 8:41 pm, Elliott, Carrie Lynn <...> wrote: > > Daniel, > > Of course! Thank you for getting back to me even though you're currently > traveling. I really appreciate it. > > I included the columns, limits, defaults, and units files for you that you > should need to reproduce the plot. I've also included the first few frames > of output that need to be read in. Since I have it handy, I'll also attach an > image of what the plot should look similar to. I'm pretty sure the plot was > rendered for output10_0001. If you could try rendering position and density, > preferably with the 6th nearest neighbor for the best comparison. > > > Best, and thank you again, > > Carrie > > > > > ________________________________________ > From: Daniel Price <[email protected]> > Sent: Thursday, June 18, 2015 3:26 PM > To: Elliott, Carrie Lynn > Subject: Re: Splash SPH software issue update - ubuntu VM > incompatibility/problem? > > Hi Carrie, > > If you can send the file you’re trying to visualise, and the exact steps > needed to reproduce the problem, I can see if there is a problem from this > end. > > Cheers, > > Daniel > > >> On 18 Jun 2015, at 7:54 pm, Elliott, Carrie Lynn <...> wrote: >> >> Daniel, >> >> After some further investigation of the problem (including but >> not limited to creating a different format of fabricated data input files, >> recompiling splash after installing all possible needed updates, checking >> the installation of cairo/X11 for errors, compiling with a debug flag and >> picking through the output log, trying a new columns file, getting someone >> to try installing a virtual machine and trying to use splash, since that is >> the method I'm using and so on) the problem is not limited to just my >> virtual machine - a friend of mine who's helping me to figure out the issue >> said he got similar problems in a ubuntu virtual machine. >> >> So, my question is now have you received any word of issues in the >> past on a ubuntu virtual machine? Is the method of a guest OS problematic >> for the plotting implementations of splash? >> >> >> Best wishes, >> >> Carrie >> >> From: Elliott, Carrie Lynn >> Sent: Thursday, June 11, 2015 4:12 PM >> To: [email protected] >> Subject: Splash SPH software issue / strange bug in rendering >> >> >> >> Hello Daniel, >> >> Thank you for your time, first and foremost, in trying to answer my >> question. This question was also posted to the user forum, so if you see it >> there, I'd prefer a direct email back instead! >> >> >> >> >> I'll try to be as specific about the problem as possible, and I'll also >> attach the interaction in the terminal as a text file, in case that would >> help you identify what's going on; hopefully you will be able to provide >> some wisdom on what is wrong or what I have done incorrectly. >> >> I'm trying to use matlab dump files generated to get some plots in splash of >> position vs. log density of a proto neutron star during core collapse. These >> dumps were generated by someone else on their computer, and directly >> transferred to me. Ideally, using his data I should render the same plots. >> We are using nearest neighbor smoothing methods to interpolate density for >> different distances for each tracer particle. >> >> I noticed during render that the warning 'WARNING NaNs in rendered quantity' >> output to the screen when plotting position vs. density frames. This to me >> seems like a clue of what's going on, but I still can't figure a solution >> out, because even in just a position vs. smoothing length (no NaN warning) >> plot, it's still coming out as one solid color when the smoothing lengths >> are NOT all the same in the data file. In the data file, density is clearly >> not uniform across the board, but should look very differentiated. >> >> >> This is what the first frame should look like (splash_0000_1.png) >> >> When I render it, it appears as the second attachment (screenshot.png) >> >> I know that my plot is not the 6th nearest neighbor density, but the same >> results happened for all column options - log density, 4th 6th and 10th >> nearest neighbor log densities, smoothing length, log radius, everything is >> rendering "blank". >> What's weird is that the limits are being read in correctly, because both >> plot's density scales are on the same order of magnitude if you check the >> images. Not exactly the same, but still on the same order. >> >> Working my way down to the simplest plot possible, position alone, even the >> particle positions aren't being plotted - what is rendered is a blank white >> plot. So something, some setting or configuration error must be going on >> with splash, at least this is my guess. >> >> Just a side note, when I back out of the directory where columns, >> splash.limits, and splash.defaults files live, the simple x-y 2D plot of >> position works. Even just without columns, it works, and all the timesteps >> also work and show the shock wave travel out through the tracers. Why, I >> cannot figure out. But I can't render the density values I need because they >> were identified by the columns file, and splash will only let me render 2 >> dimensions without it. >> >> I would be greatly appreciative of any help, advice, or any bit really of >> your opinion. >> >> >> Thanks, >> >> Carrie Elliott > > <columns><splash.defaults><splash.limits><splash.units><output10_0001><output10_0009><output10_0017><output10_0025><output10_0033><splash_0000.png> -- You received this message because you are subscribed to the Google Groups "SPLASH users forum" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/splash-users. For more options, visit https://groups.google.com/d/optout.
