Yongsop Hwang wrote: > Dear all, > > I have a problem in my harminv. > I ran a meep ctl file and got an error message as followed. > > ----------- my error message -------------- > harminv1:, frequency, imag. freq., Q, |amp|, amplitude, error > harminv: failure on line 504 of harminv.c: failed convergence in ZGEEV > ------------------------------------------- > > I got this error on both meep-mpi(cluster) and meep(single process) > calculation. > I got this error message for the metal(dispersive material) calculation but > not for the dielectric(non-dispersive material) one. > > My harminv is harminv 1.3.1. > > I ran the same ctl file on another single process server with harminv 1.3.0, > then I got results without any error. So I think my harminv 1.3.1 has a > problem. > > Here is my ctl file. > I hope someone can give me an answer. > ------------- my ctl file ---------------- > (define-param sz 5) ; size of cell in z direction > (define-param dpml 0.5) ; PML thickness > (set! geometry-lattice (make lattice (size no-size no-size sz))) > > (set! eps-averaging? false) > > (set! geometry > (list (make block (center 0 0 0) (size infinity infinity sz) > (material (make dielectric (epsilon 1.0) > (polarizations (make polarizability > (omega 1e-20) (gamma 0) (delta-epsilon 1e+40)))))) > (make block (center 0 0 0) (size infinity infinity 0.1) > (material (make dielectric (epsilon 4.0)))) > )) > > (set! pml-layers (list (make pml (direction Z) (thickness dpml)))) > (set-param! resolution 80) > > (define-param fcen 0.25) > (define-param df 5) > (set! sources (list > (make source > (src (make gaussian-src (frequency fcen) (fwidth df))) > (component Ez) (center 0 0 0)))) > > (define-param k-interp 8) > ;square lattice > (define-param k-points (list (vector3 0 0 0) ; Gamma > (vector3 2 0 0) ; X > )) > (set! k-points (interpolate k-interp k-points)) > > (run-k-points 300 k-points) > ------------------------------------------------------- > > Thanks in advance, > Yongsop Hwang > > > > _______________________________________________ > meep-discuss mailing list > [email protected] > http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss > > > Hi, Yongsop Harminv uses LAPACK subroutines to solve the corresponding eigen-problems (ZGEEV in your case). The error you get is clear: *failed convergence in ZGEEV
*You should output and view your data carefully, which may tell you what is wrong. Yours, Zheng 2008-11-13.
_______________________________________________ meep-discuss mailing list [email protected] http://ab-initio.mit.edu/cgi-bin/mailman/listinfo/meep-discuss
