changelog

采用FDTD时域有限差分法计算电磁波的传播问题等。

Sat Feb  5 12:46:16 EST 2005  [e-mail]
  * fix comment

    M ./src/energy_and_flux.cpp -3 +3

Sat Feb  5 12:24:51 EST 2005  [e-mail]
  * rename Slicing to FieldOutput (since we also count hdf5 output under this)

    M ./src/h5fields.cpp r1
    M ./src/meep.h r1
    M ./src/slices.cpp r1
    M ./src/time.cpp r1

Sat Feb  5 12:19:11 EST 2005  [e-mail]
  * more verbose output by default, with option to suppress by setting meep::quiet = true

    M ./configure.ac -10
    M ./src/bands.cpp -6 +7
    M ./src/fields.cpp -2 +3
    M ./src/h5fields.cpp +2
    M ./src/meep.h -5 +3
    M ./src/mympi.cpp -2 +4
    M ./src/output_directory.cpp -2 +3
    M ./src/step.cpp -2 +9
    M ./src/structure.cpp +3
    M ./src/time.cpp -6 +6
    M ./tests/2D_convergence.cpp +1
    M ./tests/bench.cpp +1
    M ./tests/bragg_transmission.cpp +1
    M ./tests/convergence_cyl_waveguide.cpp +1
    M ./tests/cylindrical.cpp +1
    M ./tests/flux.cpp +1
    M ./tests/h5test.cpp +1
    M ./tests/integrate.cpp +1
    M ./tests/known_results.cpp +1
    M ./tests/one_dimensional.cpp +1
    M ./tests/physical.cpp +1
    M ./tests/symmetry.cpp +1
    M ./tests/three_d.cpp +1
    M ./tests/two_dimensional.cpp +1

Sat Feb  5 11:22:51 EST 2005  [e-mail]
  * allocate all fields lazily (not just in 2d), ensure all field arrays are initialized to null, don't allocate imag. part for is_real, add missing is_real check to fields::initialize, various cleanups

    M ./src/boundaries.cpp -9 +12
    M ./src/fields.cpp -82 +61
    M ./src/initialize.cpp -44 +25
    M ./src/meep.h -3 +3
    M ./src/meep_internals.h +1
    M ./src/polarization.cpp -2 +4
    M ./src/sources.cpp -1 +5
    M ./tests/2D_convergence.cpp -1 +1
    M ./tests/h5test.cpp -3 +3
    M ./tests/known_results.cpp -1 +1
    M ./tests/symmetry.cpp -8 +8

Fri Feb  4 19:53:44 EST 2005  [e-mail]
  * major surgery on connect_the_chunks
  connect_the_chunks was:
     (a) a major performance bottleneck on multiprocessor systems
     (b) very difficult to modify, consisting of the same code
         copy-and-pasted many times
     (c) fairly fragile, with the loops impossible to reorder
         without breaking step.cc
  
  I've tried to fix all of those problems.  Now, instead of calling
  things like locate_point O(nc^2) where np is the number of chunks,
  it only computes it O(nc) times, and a lot of the duplication is
  removed.  It's also much less fragile, because the connections are
  stored per phase type (complex, -1, or +1) rather than in one
  order-sensitive array.
  
  However, this required major changes to boundaries.cpp and related code.

    M ./src/boundaries.cpp -273 +163
    M ./src/fields.cpp -25 +21
    M ./src/meep.h -10 +14
    M ./src/step.cpp -41 +58

Mon Jan 31 19:23:19 EST 2005  [e-mail]
  * added new methods dft_flux::scale_dfts, fields::require_component, and fields::remove_sources

    M ./src/dft.cpp -6 +6
    M ./src/fields.cpp -7 +22
    M ./src/meep.h -2 +8
    M ./src/sources.cpp -9 +12
    M ./tests/bragg_transmission.cpp -1 +1

Thu Jan 27 00:32:41 EST 2005  [e-mail]
  * prelimiary support for "D" sources, which multiply J by epsilon

    M ./src/sources.cpp +9

Thu Jan 27 00:30:34 EST 2005  [e-mail]
  * support symmetry-based reductions of geometric_volume_list for efficient computation e.g. of DFT fluxes etc

    M ./src/dft.cpp -45 +22
    M ./src/meep.h -4 +2
    M ./src/meep/vec.h -8 +24
    M ./src/vec.cpp +63

Thu Jan 27 00:29:49 EST 2005  [e-mail]
  * bug fix in symmetry::phase_shift for Poynting vector

    M ./src/vec.cpp -1 +3

Thu Jan 27 00:28:26 EST 2005  [e-mail]
  * force newline in abort messages if one is missing

    M ./src/mympi.cpp +1

Tue Jan 25 20:19:21 EST 2005  [e-mail]
  * immediately flush master_printf output

    M ./src/mympi.cpp -1 +1

Mon Dec  6 18:04:09 EST 2004  [e-mail]
  * comment correction

    M ./src/h5fields.cpp -2 +1
    M ./src/integrate.cpp -2 +1

Mon Dec  6 17:25:02 EST 2004  [e-mail]
  * added new derived_component type for various quantities (Poynting vector, energy density) useful to output in get_field and output_hdf5

    M ./src/h5fields.cpp +68
    M ./src/meep.h +8
    M ./src/meep/vec.h -3 +41
    M ./src/monitor.cpp +36
    M ./src/vec.cpp +42
    M ./tests/h5test.cpp -15 +24

Mon Dec  6 15:08:07 EST 2004  [e-mail]
  * consolidated min/max/abs functions in meep_internals.h

    M ./src/h5fields.cpp -5 +1
    M ./src/meep_internals.h -13 +2
    M ./src/vec.cpp -6 +1

Mon Dec  6 14:55:50 EST 2004  [e-mail]
  * whoops (only compute offsets for integration on Dielectric grid)

    M ./src/integrate.cpp -1 +2

Mon Dec  6 14:40:47 EST 2004  [e-mail]
  * fixed missing namespaces

    M ./src/update_e_from_d_update.cpp -5 +1

Mon Dec  6 14:21:53 EST 2004  [e-mail]
  * consolidated/simplified fields::output_hdf5 interface

    M ./src/h5fields.cpp -74 +54
    M ./src/meep.h -25 +8
    M ./tests/h5test.cpp -3 +3

Mon Dec  6 13:44:27 EST 2004  [e-mail]
  * add -*- C++ -*- to top of header files so that emacs will use C++ mode (even though they end with .h and not .hpp or .hh)

    M ./libctl/meep-ctl-const.h +1
    M ./libctl/meep-ctl-swig.h +1
    M ./libctl/meep-ctl.h +1
    M ./src/meep.h +1
    M ./src/meep/mympi.h +1
    M ./src/meep/vec.h +1
    M ./src/meep_internals.h +1
    M ./src/ran.h +1
    M ./src/threevec.h +1

Mon Dec  6 00:54:10 EST 2004  [e-mail]
  * rewrote hdf5 output again in terms of new generic field_function (to output an arbitrary function of the fields); also, when the Dielectric component is requested we use 3/trace(average inveps) instead of eps[i], since the former is closer to what is actually used for time-stepping

    M ./src/h5fields.cpp -107 +266
    M ./src/integrate.cpp -8 +38
    M ./src/meep.h -14 +57
    M ./src/monitor.cpp -11 +66

Sun Dec  5 14:24:12 EST 2004  [e-mail]
  * whoops, forgot to add integrate.cpp

    A ./src/integrate.cpp

Sun Dec  5 14:07:52 EST 2004  [e-mail]
  * rename fields::integrate to fields::loop_in_chunks, to reflect more generic nature; added *new* fields::integrate which *does* just do integration and makes it much much easier; rewrote energy_and_flux.cpp in terms of new integrate function

     ./src/fields_integrate.cpp -> ./src/loop_in_chunks.cpp
    M ./src/Makefile.am -6 +7
    M ./src/dft.cpp -5 +5 r1
    M ./src/energy_and_flux.cpp -231 +46 r2
    M ./src/h5fields.cpp -4 +4 r1
    M ./src/loop_in_chunks.cpp -32 +47 r1
    M ./src/meep.h -7 +21 r2
    M ./src/meep/mympi.h -1 +2
    M ./src/mympi.cpp +8
    M ./src/sources.cpp -5 +5 r1
    M ./tests/integrate.cpp -50 +27

Fri Dec  3 20:02:06 EST 2004  [e-mail]
  * faster, more easily generalized output_hdf5 based on fields::integrate
  The old version of fields::output_hdf5 pre-dated fields::integrate,
  and had home-brewed code for looping over the chunks and symmetries
  and computing the output from each.  Not only was this complicated (as
  opposed to centralizing the complexity in fields::integrate), but it
  also was based on fields::interpolate onto an arbitrary grid, which
  had the following disadvantages:
  
  * doing an arbitrary interpolation for every point in a large volume
    is very slow.  This is a problem if we want to output often.
  
  * very hard to compute functions of more than one field component,
    e.g. the energy density or the average inveps dielectric, because
    on an arbitrary grid this requires communication in general.
  
  * hard to prevent different chunks/symmetries to write slightly
    overlapping data, which produces undefined results in parallel HDF5.
  
  The new code *always* outputs on the dielectric grid, which (since it
  is a fixed interpolation scheme) is a tight loop for each chunk, and
  is guaranteed by fields::integrate to be non-overlapping.  Perhaps
  even more importantly, it will now be easy to implement output
  functions of more than one field component, since no communication is
  required for the interpolation.
  

    M ./src/h5fields.cpp -206 +174
    M ./src/meep.h -3 +3
    M ./tests/h5test.cpp -25 +36

Fri Dec  3 12:25:35 EST 2004  [e-mail]
  * whoops, fixed bug in last patch (vec min/max funcs)

    M ./src/vec.cpp +4

Fri Dec  3 00:37:08 EST 2004  [e-mail]
  * added more min/max functions for vec/ivec

    M ./src/meep/vec.h +6
    M ./src/mympi.cpp +11
    M ./src/vec.cpp -10 +30

Fri Dec  3 00:36:16 EST 2004  [e-mail]
  * added snap_unit_dims to fields::integrate

    M ./src/fields_integrate.cpp -3 +19
    M ./src/meep.h -1 +2

Fri Dec  3 00:34:09 EST 2004  [e-mail]
  * comment clarification for yee2diel_offsets

    M ./src/vec.cpp -4 +6

Thu Dec  2 17:49:54 EST 2004  [e-mail]
  * stopgap fix for PML with non-diagonal inveps...is a longer-term, "proper" fix needed???

    M ./src/structure.cpp +4

Mon Nov 29 21:37:42 EST 2004  [e-mail]
  * added h5file::prevent_deadlock() for deadlock-prevention hacks in exclusive-access HDF5 mode (maybe we should just ditch support for that and require HDF5 to be compiled with MPI?)

    M ./src/dft.cpp +2
    M ./src/h5file.cpp -4 +10
    M ./src/meep.h +1
    M ./tests/bragg_transmission.cpp +2

Mon Nov 29 19:28:53 EST 2004  [e-mail]
  * completely revamped structure and structure_chunk initialization
  In the previous version of the code, the typical initialization
  sequence went like this:
  
  1) create structure (initializes chunks, evaluates epsilon)
  2) set PML (re-initializes chunks, shuffles epsilon around)
  
  Step (2) is problematic because the communication of epsilon to the
  new chunks was performed point-by-point.  Not only is this incredibly
  slow, but it actually causes Tungsten's MPI implementation to crash
  (probably because there are too many messages).  Moreover, it is
  completely unnecessary.
  
  In the new version, the PML boundaries are specified when the structure
  is first created, so that the chunks need only be computed once and are
  never re-shuffled around.
  
  INCOMPATIBLE API CHANGE:
  
    old:
       structure s(vol, eps, splitting, symmetries);
       s.use_pml_everywhere(1.0);
  
    new:
       structure s(vol, eps, pml(1.0), symmetries, splitting);
  
  In general, the PML boundary regions are specified by *summing* terms
  of the form pml(thickness, direction, High/Low).  If one or both of
  the last two arguments is omitted, it means *all* sides and *all*
  directions, respectively.  So, a more complicated example is:
  
    old:
       structure s(vol, eps, splitting, symmetries);
       s.use_pml(X, High, 1.0);
       s.use_pml(X, Low, 1.0);
       s.use_pml(Y, High, 1.0);
  
    new:
       structure s(vol, eps, pml(1.0, X) + pml(1.0, Y, High), 
                   symmetries, splitting);
  
  To specify *no* PML, you use:
     
     structure s(vol, eps, no_pml(), symmetries, splitting);
  
  This is actually the default, if you omit the boundary-region argument
  along with splitting (default = 0) and symmetries (default =
  identity).
  
  INCOMPATIBLE API CHANGE:
  
    old:
       s.set_epsilon(eps, minvol, do_aniso_averaging);
  
    new:
       s.set_epsilon(eps, do_aniso_averaging, minvol);
  
  The default for the last two arguments is true and 0, respectively, so
  in the usual case where you are using this function because you want
  to turn on averaging (currently not the default), you just do:
  
     s.set_epsilon(eps);

    M ./doc/complicated.tex -2 +1
    M ./doc/epsilon_polariton_1d.tex -2 +1
    M ./doc/lossgain_epsilon.tex -5 +4
    M ./doc/nonlinear.tex -5 +3
    M ./doc/omniguide.tex -1 +1
    M ./doc/simple.tex -3 +2
    M ./src/meep.h -31 +86
    M ./src/structure.cpp -279 +139
    M ./tests/2D_convergence.cpp -1 +1
    M ./tests/bench.cpp -3 +2
    M ./tests/bragg_transmission.cpp -4 +2
    M ./tests/convergence_cyl_waveguide.cpp -3 +3
    M ./tests/cylindrical.cpp -10 +8
    M ./tests/flux.cpp -12 +6
    M ./tests/h5test.cpp -3 +3
    M ./tests/integrate.cpp -1 +1
    M ./tests/known_results.cpp -5 +2
    M ./tests/one_dimensional.cpp -4 +4
    M ./tests/physical.cpp -4 +2
    M ./tests/symmetry.cpp -42 +38
    M ./tests/three_d.cpp -10 +7
    M ./tests/two_dimensional.cpp -22 +12

Sun Nov 28 19:01:35 EST 2004  [e-mail]
  * provide high-resolution wall_time() function in lieu of clock().  Automatically print elapsed time upon program completion.  Don't use exit(0) in main()...use return 0 so that ~initialize() is called.

    M ./configure.ac +7
    M ./src/fields.cpp -2 +2
    M ./src/meep.h -1 +1
    M ./src/meep/mympi.h +5
    M ./src/mympi.cpp +31
    M ./src/time.cpp -10 +8
    M ./tests/bench.cpp -20 +19
    M ./tests/cylindrical.cpp -1 +1
    M ./tests/flux.cpp -1 +1
    M ./tests/one_dimensional.cpp -1 +1
    M ./tests/three_d.cpp -1 +1
    M ./tests/two_dimensional.cpp -1 +1

Wed Nov 24 20:04:09 EST 2004  [e-mail]
  * add stride0 function that returns 0, instead of 1, for unused dimensions, so that we can easily distinguish them from "real" dimensions in macros etc.

    M ./src/meep/vec.h -7 +9
    M ./src/vec.cpp -5 +6

Wed Nov 24 20:03:39 EST 2004  [e-mail]
  * don't use make_average_eps in omniguide.tex

    M ./doc/omniguide.tex -1 +2

Wed Nov 24 17:49:55 EST 2004  [e-mail]
  * remove structure::make_average_eps (apparently unused, non-functional, pointless function)

    M ./src/meep.h -2
    M ./src/structure.cpp -20

Fri Nov  5 12:45:13 EST 2004  [e-mail]
  * updated acx_mpi.m4

    M ./m4/acx_mpi.m4 -4 +4

Fri Oct 29 21:30:13 EDT 2004  [e-mail]
  * added another joke