Resolve "Enabling free Maxwell propagation in multisystem mode, part 2/4"
Merged
requested to merge 206-enabling-free-maxwell-propagation-in-multisystem-mode-part-2-4 into develop
Description
Allow Maxwell propagation with external currents.
News snippet
Allow Maxwell propagation with external currents.
Checklist
-
I have checked that my code follows the Octopus coding standards -
I have added tests for all the new features added in this request.
Closes #206 (closed)
Edited by Franco Bonafé
Merge request reports
Activity
added 210 commits
- 35b97d62 - defining exchange energy
- 83b59b8d - scf.f90 file: fixing restart, adding output of correlator, pt_number, pt_exchange, ss
- ef3117d1 - update Makefile.am to deal with new files photon_mode.F90 and xc_oep_qed_inc.F90
- 870990f0 - adding pt_exchange to energy_calc.F90
- 04cdbbbb - adding pt_exchange to ks%calc_energy
- 78f217fc - photon kli by Christian Schaefer (christian.schaefer@mpsd.mpg.de)
- 052754c6 - photon oep
- 5bcaac41 - photon oep _inc file
- fa801020 - two new files, photon mode and xc_oep_qed_inc.F90
- dc716e01 - bugix
- adb5da25 - adding correct output of normss
- b21afa32 - adding line break for cuda/mpi compiler
- 0c4f30a5 - comment with reference to photon oep paper on arxiv
- 0d41cd74 - adding experimental warning if photons=yes
- 3d10dfc6 - Merge branch 'photon-oep' into 'master'
- 76b77030 - Merge branch 'master' into 'photon-oep'
- 614b5f92 - resolving two problems due to merge with master branch (v8.2)
- ce332a1e - Merge branch 'photon-oep' into 'photon-oep'
- a7145076 - Replace Makefile.am by Christian Schaefer
- 00fe9134 - Revert "Replace Makefile.am by Christian Schaefer"
- 57156522 - Merge branch 'revert-a7145076' into 'photon-oep'
- f39446a6 - changed oep to many photon mode syntax
- efd2a538 - removing comments
- f79322f4 - updating the photonmode file
- e2802f40 - fixing problems for spinresolved calculations
- 8d4bf07f - Merge branch 'develop' into HEAD
- bad9242d - using nst
- 9b707c9f - oep should only run over occupied states
- eb27abc4 - adding a line that has been removed unintentionally
- 0b2aad69 - integer need to be declared
- 28d26828 - spin unrestricted OEP should work now, we need some proper tests
- d9dd8c0a - spin-resolved related changes
- 0282905d - fixing the output of the photon_correlator
- 1ec557a3 - Merge branch 'develop' into 'photon-oep'
- 8f934696 - removing comments that were not up to date
- be143cc4 - changing order of deallocation, photonmode has only to be called if run with photons
- b5bfd3a0 - Merge branch 'develop' into 'photon-oep'
- c190a81c - fixing the intent argument for KLI routine
- 3cf80565 - Included center-of-charge shift for the dipole operator into the photon
- dc659f12 - Included KLIpt_coc parameter (default=false) to enable switching between
- 00c76a19 - Merge branch 'develop' into HEAD
- 3c1a09bf - fixing the errors of the testsuite, now everything should run through
- 377db06c - Merge branch 'photon-oep' into 'develop'
- 6a79d4d4 - Merge branch 'develop' into 'photon-oep'
- 6b18189a - include break-line in coc-shift to avoid Fortran maxcharacter issue
- 9e191c0a - fixing the cuda mpi error by explicitely specifying dimensions
- aa38ac74 - fix reference and only allow spin-unpolarized (for now)
- 95b324c6 - using axpy for increased performance
- af98324b - Adding the support for complex-to-complex ffts.
- 6197b3f3 - Merge branch 'develop' into complex_fft
- 73012ec6 - The complex FFT caanot be used with dressed Poisson solver.
- 18dde712 - If we want to do complex-to-complex FFTs, we need to have memory allocated to the proper size.
- efc8bac6 - The two-body integrals and related exchange integrals are now computed using a...
- a04bcff8 - Merge branch 'develop' into photon-oep
- 6fd314ce - adding the namespace variable for the linear solver
- f4060f64 - adding a testfile for photon OEP and KLI
- 4e6a1dcc - changing output pt_exchange/number -> photon exchange/number
- d3b96e9e - adding the test input files
- 17366213 - Explicitly specify dimension
- f3a8abf8 - also print photon energy and number for KLI
- a2f626e4 - Adding message not implemented for Slater approximation and one fixing typo.
- 51c21439 - Some small fixes related to coding standards.
- e5eb9049 - It is now possible to run HF and hybrid calculations for solids for an...
- f0500041 - Merge branch 'develop' of gitlab.com:octopus-code/octopus into develop
- 087581f4 - In case FFT kernel is real, we cannot use the complex-to-complex routine from zpoisson_fft_solve.
- efb49332 - Fix a warning.
- fb61e929 - Fix the behavior for the calculation of the Coulomb singularity in...
- d8238bee - Fix the use of singularity for screened hybrids in isolated systems.
- 0a506208 - first set of cleaning of code as suggested by Micael
- ba1369c7 - fixing loop
- 3ac35680 - changes to comply with coding standards
- 62fae345 - coding standard, cleaning photon mode module, moving lr to oep module, using dotu in mf_dotp
- 9d22f3da - Merge branch 'develop' into photon-oep
- 80e9659f - Merge branch 'develop' into photon-oep
- a98b5d53 - Fix a problem with the input variable FFTPreparePlan, which was not working.
- fc8c2de9 - Fix merge conflict with develop.
- 7e727119 - Merging with develop.
- 9c87b457 - Proper comparison of logicals.
- 1d6fbc4c - Fix a left over from merge conflict.
- 0f4d6dc4 - Proper comparison of logicals.
- 0e6227f8 - Add a missing change for ISF.
- 759d1056 - The exhange operator is only initialized when needed.
- 295220a1 - RDMFT also need the exchange operator.
- 016106aa - We also need it for the OEP.
- cfea1133 - Several improvements to the linked lists and how the systems are handled.
- 6e707143 - Merge branch 'improve_system_linked_list' into 'develop'
- 9896aeba - First step towards implemented a new logic for propagators and multisystems.
- 35065f31 - Adding a demonstrator based on celestial_bodies. Not working yet.
- bf45823b - Making the celestial dynamics more complete, with a implementation of the...
- 7e5c7065 - Small changes to make the celestial dynamics produce a meaningful result.
- 98b2b4e3 - Celestial body's properties are now parsed from the input file.
- bb36e0d3 - Add a namespace to the abstract system, as all systems should have their own namespace.
- 969e7d16 - Improve class arguments naming scheme consistency. Minor cosmetic changes.
- 2dd4d59d - Make sure forces are initialized to zero.
- d2ebb002 - No need to have a init deferred method for propagators.
- fa5f72de - The propagator verlet is now initialized with a function that overloads the type.
- 29d62250 - Move the calculation of the force between two celestial bodies to a new...
- 7a7aa5db - Make propagator_finished a method of the abstract propagator. Now only...
- dc6c2de7 - Print some debug information when running a propagation step.
- d6c829d0 - Nicer output during the celestial dynamics test mode. Systems are now required...
- 6ed63fa3 - Actually stop the propagation once we are finished!
- 1c584981 - Add routines to finalize celestial body.
- ff3dc7eb - celestial_body_init is not a class method anymore. Instead, it is now a...
- 54325e8b - The multisystem framework is now aware of the existence of celestial bodies.
- 3f762d04 - Collapse two if statements.
- 5af35405 - Merge branch 'photon-oep' into 'develop'
- 129e256f - Major refactoring of the new framework for propagators and interactions.
- 05fe5fa3 - Add new CelestialDynamics.html file to Makefile.
- 95506db6 - New methods for the celestial body class to write information during the td run.
- 5bdaccad - Add units of the celestial body's input options.
- 521aeede - Make the propagator a component of the abstract system.
- 23e47c80 - Update celestial dynamics test.
- 752db736 - Moved the known propagator operations to propagator_abst_oct_m.
- 0a282852 - Renamed system_dt to system_dt_operation and the corresponding class method to dt_operation.
- 61c9e06e - Move the set_propagator method to the abstract system.
- 6121b197 - Update celestial dynamics test reference values and tolerances.
- b07514d5 - Merge branch 'toward_multisystem_propagators' into 'develop'
- 9f8e4af4 - Add a missing initialization.
- 934f033f - Fix a bug with dim<3 and cell angles.
- 75a142f3 - Combining the routines zpoisson_solve_fft and dpoisson_solve_fft to avoid code duplications.
- 82401438 - Fix bug: wrong arrays were passed
- c44cb917 - Refactor allocation code and fix memory leak for deallocation
- c6cbfe42 - Remove gpg output for the git revision that is baked into the octopus binary.
- 682e309b - This value should always be initialized.
- 1f764cf1 - Adding first very simple simulation clock implementation
- 49b2c7bd - Add granularity to the clock to be able to compare clocks globally.
- 5b026b3a - Add number of algorithmic steps to the abstract propagator
- f0797dd3 - Rename routines for clock comparison.
- 1463f979 - Add function to decrement clock by one tick.
- 01a7d278 - Adding function is_later_with_step to clock.
- 43694eeb - Overload operators for clock.
- b89b4dd8 - Add print function
- a2947f71 - Update overloaded operator definition and epoch timers
- 1886d97b - Add namespace to clock and improve output format of clock print function
- f465ed1e - Add unit test for clock
- 71533e3c - Switch from boolean to integer output values for the testsuite
- aad1e0df - Add regression test for clock
- 3fb17432 - Add finest system wide tick to clock print function
- 3f16abbf - Add more print statements to clock unit test
- b652cf86 - Break long lines for the gfortran configuration in the buildbot
- 615b80d0 - Remove unused variables
- 46c14fd3 - Revert to test_celestial_dynamics routine from develop
- 0b1bbb07 - Merge branch 'complex_fft' into 'develop'
- 26916603 - Merging with develop. This does not compile for the moment.
- 16ac056d - Merge branch '276-Add-support-for-gpg-signatures-in-git-commits' into 'develop'
- d6619f33 - Merge branch 'improved_fftw_support' into 'develop'
- 4246000f - Merge branch 'fix_cell_angle_low_dim' into 'develop'
- d4d507e1 - Addressing issues raised in the code review
- 9597252a - Further updates based on the code review for the MR
- c9c1b874 - Allow to increment or decrement the clock ticks by more than one step
- 65a0579d - Rename simulation_clock to clock
- ed6be314 - Remove epoch variables from clock
- 3e2e81c0 - modules and variables renamed to psolver_
- 73f839ae - Add documentation for the clock
- 81938a0f - reverted to HAVE_LIBIFS to avoid breaking the tests for other branches
- 17cbaa1b - Merge branch '273-add-simulation-clocks-for-multi-system-synchronization-v2' into 'develop'
- 033cafb9 - Make batched current default for non-orthogonal cells
- db0b7650 - Remove unnecessary application of the phase to the gradient.
- 56c3efd3 - resolve style issues
- f011386d - Fix a memory leak in the current calculation in GPU.
- 9dd277c6 - Adding an assert to check the assumption used in the code.
- f0ea83f6 - Split the set function of the clock to a full set, including namespace and a time set only
- 688aa79e - Adding clock print_str and print_message functions
- 1efd8bdd - Merge branch '278-rename-poisson_libisf_oct-to-poisson_psolver_oct' into 'develop'
- aaf12cb7 - Align intents and remove one trim
- a9c2efb3 - Rename clock_set_all to clock_copy
- 9cdd1cdd - The kernel of the FFT is now isolated from the Poisson solver itself. This...
- b1479646 - Fix some problems with the previous commit and adjust a test, as the code was...
- c2479926 - Merge branch 'current_batch_gradient' into 'develop'
- 34ba4a64 - Only compile Libyaml if LibISF was not found.
- 4cef4fa6 - Add check for latest version of PSolver.
- 30330c57 - PSolver support updated to allow for new and old API's. New API adds support...
- ae25c505 - Renamed libisf to psolver in several places (documentation, variables, tests, etc)
- 75382e44 - Mark use of olver versions of PSOlver as deprecated.
- aadcbf64 - Rename poisson_libisf.F90 to poisson_psolver.F90.
- 4f65635d - warning fixed (hartree_energy)
- 39cc4cfe - Merge branch '279-update-poisson_psolver_oct-to-allow-for-both-the-old-and-new-api' into 'develop'
- 3babc8b7 - Simplifies the logic of the code.
- eb012bb3 - Remove unused first method and generic add method from linked list.
- ec587d80 - Add linked list method to iterate
- 77db56aa - Rename functions in linked list module to make them less generic and closer to...
- e493c6ed - Remove namespace from clock
- 70d50756 - Remove print_message function from clock
- 1e03e53d - added explicit specification of MPI group for the Psolver
- 864c8766 - Merge branch 'develop' into exchange_for_solids
- e216954e - Use new linked list iterate method to simplify some loops.
- 875eca83 - Merge branch 'develop' into 280-improved-clock-printing-and-setting
- 4ea5a651 - Since the propagator is a list of instruction, we make it an extension of a linked list.
- 51b48311 - Removed iterator from linked list.
- 0369ff53 - Add wrapper class for list of interactions.
- 128c9bac - Adding a nullify routine to avoid problem in releasing memory hat was not allocated.
- f7a50bb7 - For some strange reason this value got modified in the commit...
- 38dbcd3d - Use messages_fatal instead of output to stderr
- 1d954d96 - Add a list_counter class that implements the necessary methods to iterate over...
- d28dc311 - The propagator now uses the list counter.
- 6ffa8fc3 - Renamed counter_end_of_list to counter_has_next.
- ceeaf75a - Add missing push/pop
- e85f44cb - Transform the list counter into a proper iterator.
- 75233c5f - enabled PSOLVER for serial builds
- d2e3b22d - Merge branch 'more_lists' into 'develop'
- 2045947a - The range of the soft Coulomb potential is now set by according to...
- c9b1fb66 - Merge branch '280-improved-clock-printing-and-setting' into 'develop'
- a3d54419 - Merge branch 'soft_coulomb_threshold' into 'develop'
- cf5ea6ea - The clock-related operator and the assignment are now methods of the class....
- f1f76808 - added some comments
- 7f188017 - Merge branch 'clock_class_operators' into 'develop'
- 17338809 - Merge branch 'exchange_for_solids' into 'develop'
- 58eb303d - Merge branch '282-enable-psolver-for-sequential-runs' into 'develop'
- e18a00c0 - Fix a problem with array dimensions.
- 39263646 - Merging with develop and fixing the conflicts.
Toggle commit listCodecov Report
Merging #784 into develop will increase coverage by
1.51%
. The diff coverage is54.34%
.@@ Coverage Diff @@ ## develop #784 +/- ## =========================================== + Coverage 69.26% 70.78% +1.51% =========================================== Files 505 500 -5 Lines 98106 97621 -485 =========================================== + Hits 67958 69104 +1146 + Misses 30148 28517 -1631
Impacted Files Coverage Δ src/grid/mesh.F90 70.25% <ø> (-1.86%)
src/grid/restart.F90 89.00% <ø> (-2.49%)
src/hamiltonian/hamiltonian_mxll.F90 21.64% <ø> (+2.49%)
src/math/maxwell_function.F90 25.42% <ø> (-0.28%)
src/states/states_mxll.F90 51.44% <ø> (-7.31%)
src/system/system_abst.F90 67.85% <ø> (-4.40%)
src/td/propagator_base.F90 0.00% <ø> (ø)
src/td/propagator_mxll.F90 31.95% <ø> (+10.87%)
src/td/td.F90 75.07% <ø> (+0.40%)
src/td/td_write.F90 58.18% <ø> (-0.21%)
... and 272 more
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update 32bd0e9...32bd0e9. Read the comment docs.Edited by Codecovadded 1 commit
- b456d9b3 - Add a missing allocation of the pointer inside the system_mxll_init routine.
added 634 commits
-
c392a77b...32bd0e92 - 634 commits from branch
205-enabling-free-maxwell-propagation-in-multisystem-mode-part-1-4
-
c392a77b...32bd0e92 - 634 commits from branch
added 8 commits
-
32bd0e92...ef1b008a - 5 commits from branch
develop
- f5c39894 - Add necessary calls inside mxll_propagation_step and transform_rs_densities...
- f29ad512 - Add external current test files, still doesn't match with previous tests but numbers are similar
- e65c50e5 - Add message_fatal for Maxwell-matter coupling
Toggle commit list-
32bd0e92...ef1b008a - 5 commits from branch
assigned to @fbonafe
assigned to @heiko.appel and unassigned @fbonafe
added 1 commit
- 2bfc2e75 - Fix calls in mxll_propagation_step and get_rs_density_ext and update test...
changed milestone to %11.0
added 1 commit
- 8fc867b0 - Change external current pointers to allocatables in states_mxll
- Resolved by Micael Oliveira
Please register or sign in to reply