[Octopus-notify] svn commit: r7532 - in trunk: src/scf testsuite/finite_systems_1d testsuite/finite_systems_2d testsuite/finite_systems_3d testsuite/linear_response testsuite/optimal_control testsuite/periodic_systems testsuite/real_time by dstrubbe

[octopus-svn at tddft.org]

      Author: dstrubbe
        Date: Thu Mar  3 06:42:10 2011
New Revision: 7532
   Changeset: http://www.tddft.org/trac/octopus/changeset/7532

Modified:
   trunk/src/scf/scf.F90
   trunk/testsuite/finite_systems_1d/09-morse.01-gs.inp
   trunk/testsuite/finite_systems_1d/09-morse.02-unocc.inp
   trunk/testsuite/finite_systems_2d/05-spin_precession.01-ground_state.inp
   trunk/testsuite/finite_systems_2d/07-modelmb.01-gs.inp
   trunk/testsuite/finite_systems_3d/01-carbon_atom.01-ground_state.inp
   trunk/testsuite/finite_systems_3d/01-carbon_atom.02-ground_state.inp
   trunk/testsuite/finite_systems_3d/01-carbon_atom.03-ground_state_double_grid.inp
   trunk/testsuite/finite_systems_3d/02-curvilinear_coordinates.01-gygi.inp
   trunk/testsuite/finite_systems_3d/04-jellium.01-ground_state.inp
   trunk/testsuite/finite_systems_3d/05-output.test
   trunk/testsuite/finite_systems_3d/06-restart.01-ground_state.inp
   trunk/testsuite/finite_systems_3d/10-fullerene.01-gs.inp
   trunk/testsuite/finite_systems_3d/12-forces.01-N2_gs.inp
   trunk/testsuite/finite_systems_3d/12-forces.02-N2_gs.inp
   trunk/testsuite/finite_systems_3d/12-forces.03-N2_gs.inp
   trunk/testsuite/finite_systems_3d/18-current.01-current.inp
   trunk/testsuite/finite_systems_3d/20-eigensolver.05-evolution.inp
   trunk/testsuite/finite_systems_3d/20-eigensolver.test
   trunk/testsuite/finite_systems_3d/21-lobpcg.02-spinors.inp
   trunk/testsuite/finite_systems_3d/22-time_propagation.test
   trunk/testsuite/functionals/02-xc_2d.00-hartree.inp
   trunk/testsuite/functionals/02-xc_2d.01-hf.inp
   trunk/testsuite/functionals/04-oep.01-jellium-lsda.inp
   trunk/testsuite/functionals/04-oep.02-jellium-exx_kli.inp
   trunk/testsuite/functionals/04-oep.03-jellium-full_exx.inp
   trunk/testsuite/linear_response/02-h2o_pol_lr.01_h2o_gs.inp
   trunk/testsuite/linear_response/05-polarizability.03-finite-diff.inp
   trunk/testsuite/optimal_control/06-zbr98.01-ground_state.inp
   trunk/testsuite/periodic_systems/02-cosine_potential.01-ground_state.inp
   trunk/testsuite/periodic_systems/02-cosine_potential.02-unocc.inp
   trunk/testsuite/periodic_systems/06-h2o_pol_lr.01_gs.inp
   trunk/testsuite/periodic_systems/08-diamond.01-gs.inp
   trunk/testsuite/real_time/11-tdmagnetic.02-unocc.inp

Log:
* Removed the spurious sqrt in the definition of abs_dens. Now it is \int | rho_in - rho_out | d3r as intended in r3856, so that the code now agrees with the documentation.
* rel_dens = abs_dens / qtot is of course affected too. New rel_dens = qtot * old^2. I reset the default ConvRelDens to a value that is identical if qtot = 10.
* Changed all ConvAbsDens in testsuite to the square of the previous number, so the convergence will be equivalent.
* Changed most ConvRelDens to the corresponding new value. A few I left alone that did not make a test fail.
* Updated test values for finite_systems_3d/05-output, 20-eigensolver, and 22-time_propagation for the new default ConvRelDens.
* Added same ConvRelDens to finite_systems_3d/01-carbon_atom.03-ground_state_double_grid.inp as in 01 and 02 input files.

The 08-diamond test is failing, but was before this change. Not sure why.
Also, many of the tests complain that pseudo radii are too large.