Ah sorry, I thought it was MPB, not MEEP. Well based on the wiki, I would guess setting "eps-averaging?" to false is is the equivalent for MEEP:
"eps-averaging? [boolean] If true (the default), then subpixel averaging is used when initializing the dielectric function (see the Farjadpour et al. reference in Citing Meep). The input variables subpixel-maxeval (default 100000) and subpixel-tol (default 1.0e-4) specify the maximum number of function evaluations and the integration tolerance for subpixel averaging. Increasing/decreasing these, respectively, will cause a more accurate (but slower) computation of the average ε (with diminishing returns for the actual FDTD error)." 2013/11/5 Niederberger, Armand <armand.niederber...@hp.com> > > Thanks, Mike! > What version of meep are you using? > I'm using some cluster version; may this be the problem? > > In fact, I get error messages when using your file (below). > Furthermore, when I just introduce your mesh-size line into my control file, > I get > ERROR: Unbound variable: mesh-size. > Cheers > Armand > > Using MPI version 2.2, 12 processes > ERROR: Unbound variable: num-bands > ERROR: Unbound variable: num-bands > ERROR: Unbound variable: num-bands > ERROR: Unbound variable: num-bands > ERROR: Unbound variable: num-bands > ERROR: Unbound variable: num-bands > MPI Application rank 4 exited before MPI_Finalize() with status 1 > ERROR: Unbound variable: num-bands > ERROR: Unbound variable: num-bands > ERROR: Unbound variable: num-bands > ERROR: Unbound variable: num-bands > ERROR: Unbound variable: num-bands > ERROR: Unbound variable: num-bands > > > > -----Original Message----- > From: Mike Taverne [mailto:mike.tave...@bristol.ac.uk] > Sent: Tuesday, November 05, 2013 1:08 PM > To: Niederberger, Armand; meep-discuss@ab-initio.mit.edu > Subject: Re: [Meep-discuss] epsilon-input-file problem > > Try setting mesh-size to 1. If it is different, MPB will average the epsilon > values over a mesh of "mesh-size" points around each grid point. > > I'm attaching a test file I just created. > I didn't compare any simulation results, just the created .h5 files. > > I don't fully understand the mesh-size averaging, but setting it to 1 seems > to work to get identical epsilon outputs. :) > > > On 05/11/13 20:17, Niederberger, Armand wrote: > > Dear all, > > Have you successfully used the epsilon-input-file command in your control > > files? > > > > In my simulation of optical gratings, there seems to be a problem whenever > > I use epsilon-input-file. > > > > Concretely: > > > > - First, I simulate (normally, no epsilon-input-file) my grating > > and export the eps file > > > > - Second, I modify the control file to import the previously > > generated epsilon-input-file and delete any all geometry commands > > > > I notice the following: > > > > - When importing the epsilon file and outputting it again, the > > values changed (averaged? Scaled?) even if I am setting eps-averaging off. > > > > - The simulation no longer yields the same result if I import the > > epsilon file (e.g. as if the wavelength had changed). > > > > My questions: > > > > - Are there any parameters I have to set along with > > eps-import-file? (since re-importing does not work) > > > > - Does the order of the commands play a crucial role? (just > > checking...) > > > > N.B. I'll be happy to reproduce the code if needed but I don't like to spam > > people :) > > > > Thanks a lot > > Armand > > > > > > > > > > _______________________________________________ > > meep-discuss mailing list > > meep-discuss@ab-initio.mit.edu > > http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss > > > _______________________________________________ meep-discuss mailing list meep-discuss@ab-initio.mit.edu http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
