portelli/Grid
1
0
Fork 0
mirror of https://github.com/paboyle/Grid.git synced 2026-04-10 05:56:09 +01:00
Code Releases Activity

Commit Graph

  • 92def28bd3 Update README.md Peter Boyle 2021-06-06 04:52:05 -04:00
  • ca10bfa1c7 removing Travis CI constantly failing due to overtime (no way we can compile Grid on free time anymore) Antonin Portelli 2021-06-04 11:12:22 +01:00
  • 298a6ec51e Merge pull request #357 from mmphys/bugfix/ragged Antonin Portelli 2021-06-04 10:34:46 +01:00
  • 1cc706b2f4 Better precision Quadro 2021-06-03 17:27:49 -04:00
  • c3b99de33f In EOFA pseudofermion action, implemented M^{-1} (this costs the same as M for EOFA!) Added tests/solver/Test_eofa_inv.cc to test the above In MDWF+ID GPBC binary, tests of RHMC approx for the action / MD approxs can be performed separately using a cmdline toggle Christopher Kelly 2021-06-03 11:11:14 -04:00
  • e5dbe488a6 Merge branch 'develop' into bugfix/ragged Michael Marshall 2021-06-03 08:25:56 +01:00
  • 0e27e3847d Remove synch Peter Boyle 2021-06-03 04:24:19 +00:00
  • 9e51fa5681 Mixed precision changes Quadro 2021-06-01 13:41:02 -04:00
  • 9164cfbfc6 Mixed precision changes Quadro 2021-06-01 13:39:33 -04:00
  • 2f3a96e5de Mixed precision changes Quadro 2021-06-01 13:38:00 -04:00
  • 26aa89cb0c HMC logging Quadro 2021-06-01 13:36:17 -04:00
  • 426d2365d1 Schur factored matrix Quadro 2021-06-01 13:35:38 -04:00
  • 81bbd4e4ce Force logging improvements Quadro 2021-06-01 13:34:43 -04:00
  • b83bbe2dd1 Mixed precision Quadro 2021-06-01 13:34:05 -04:00
  • 3f2d8eb727 Mixed precision change Quadro 2021-06-01 13:33:18 -04:00
  • 60f9bf69cd These are not acccessible from device Quadro 2021-06-01 13:32:34 -04:00
  • 38d8cd228e Reusable mixed precision wrapper Quadro 2021-06-01 13:31:18 -04:00
  • 4e1e242025 Short term hack while testing force sizes Quadro 2021-06-01 13:30:42 -04:00
  • 20c6f19bb2 Virtual fix Quadro 2021-06-01 13:29:44 -04:00
  • 036270a0c5 Bug fix must preserve "virtual" Quadro 2021-06-01 13:25:44 -04:00
  • 6c506601c5 Logging the forces each trajectory useful Quadro 2021-06-01 13:25:03 -04:00
  • e1a02bb80a Added main program to reproduce 32ID ensemble with 240MeV pions and GPBC Allowed EOFA to accept different solvers for the L and R operations in the heatbath step Fixed EOFA Meofa operating on member Phi rather than input field Added derived EOFA pseudofermion variant that allows for mixed prec CG to be used in the heatbath Added forces/Test_mobius_gparity_eofa_mixed testing the above reproduces the regular EOFA To Test_gamma, added checks for the various properties of the charge conjugation matrix C=-gamma2*gamma4 in Grid basis Christopher Kelly 2021-06-01 11:44:34 -04:00
  • 393727b93b Documentation update (briefly) covering serialisation changes. For review Michael Marshall 2021-06-01 15:49:37 +01:00
  • 2b1fcd78c3 Fixes post review with Peter: a) Correct bug in isRegularShape - detect 3d matrix where 1st slice is 2x2 and second slice is 2x1; b) Synchronisation of EigenResizeCounter done by checking we're the OMP primary thread; c) Move definition of EigenResizeCounter to new file, BaseIO.cc Michael Marshall 2021-05-31 22:24:54 +01:00
  • 0a4e0b49a0 BaseIO: Added "EigenResizeCounter" to keep track of any allocations/deallocations to Eigen tensors during readback. On read, if the tensor is resized, EigenResizeCounter += delta memory (in bytes) Michael Marshall 2021-05-31 12:49:56 +01:00
  • 76af169f05 Add global namespace to Writer<T> and Reader<T> inside GRID_SERIALIZABLE_CLASS_MEMBERS (so that "using Grid" not necessary). Fix issue with output of Grid::iMatrix so that M<3>{{148,149,150,} {151,152,153,} {154155156}} becomes M<3>{{148,149,150} {151,152,153} {154,155,156}} Michael Marshall 2021-05-31 08:43:02 +01:00
  • 7b89232251 Extended HDF5 serialisation of std::vector<T> where T now also includes Grid scalar/vector/matrix Changed VectorUtils element traits to is_flattenable, because: a) contract changed on what it does; and b) no other Grid dependencies on element. Needs review. Initial tests work ... needs proper regression testing. Michael Marshall 2021-05-30 20:27:53 +01:00
  • b5aeae526f Make Cshift fields static to avoid repeated reallocaate overhead Peter Boyle 2021-03-29 21:44:14 +02:00
  • e4ff4c902a 2f Mixed precision DDHMC running and conserving hamiltonian Quadro 2021-05-27 17:10:04 -04:00
  • 2e4d4625b6 correct multinode Quadro 2021-05-25 21:08:50 -04:00
  • 10f2c2530b Temporary debug / display force structure Quadro 2021-05-25 21:08:09 -04:00
  • 375e0698dc 2 deep cleaner coding Quadro 2021-05-25 19:51:52 -04:00
  • 64b3b37476 Force test - try degenerate ratio and check for zero force Quadro 2021-05-25 18:27:46 -04:00
  • 59e1a9be4e Locally periodic option Quadro 2021-05-25 18:25:51 -04:00
  • aac1736617 Double or single Quadro 2021-05-25 18:21:48 -04:00
  • 60f814733d Could maybe simplify and require theh params always containn locally_periodic Quadro 2021-05-25 18:18:15 -04:00
  • 612e9a178a Should maybe require the params always contain locally_periodic Quadro 2021-05-25 18:17:31 -04:00
  • 21af9cf83c GetDoubledGaugeField Quadro 2021-05-25 18:16:37 -04:00
  • 3b8cb929d6 Doubled GaugeField option Quadro 2021-05-25 18:15:47 -04:00
  • 5d3046eae8 Locally periodic override option Quadro 2021-05-25 18:15:17 -04:00
  • d73063682e GetDoubledGaugeField Quadro 2021-05-25 18:14:55 -04:00
  • 59584b6605 GetDoubledGaugeField Quadro 2021-05-25 18:14:34 -04:00
  • f6d7188615 GetDoubledGaugeField useful for mixed precision or generally retrieving the double stored gauge field Quadro 2021-05-25 18:13:37 -04:00
  • b810b6f6bd Dirichlet working multinode now Quadro 2021-05-25 18:13:01 -04:00
  • e0a92dff32 update action params Quadro 2021-05-25 18:12:11 -04:00
  • ef0ddd5d04 std::vector serialisation in hdf5 uses a different format if the vector is ragged. When reading back std::vector we need to check which format we're reading (since we don't know a priori) and this involves looking for attributes that may not exist. The c++ API: a) throws; and b) prints voluminous logging. Switched to non-throwing, non-logging, C version of the API after code review. Michael Marshall 2021-05-24 18:43:55 +01:00
  • 9b73dacf50 First row might still be ragged if multi dimensional. attrExists() doesn't throw, but easier to wrap in try ... catch than to explain in comment. Michael Marshall 2021-05-22 04:34:32 +01:00
  • 244b4aa07f Serialise std::vector of numeric types as multidimensional object if size is regular ... or individually if ragged Michael Marshall 2021-05-21 20:08:56 +01:00
  • e125f0f738 Running with small dH aand Luscher's gauge link freezing scheme Quadro 2021-05-18 16:38:07 -04:00
  • 5f081d87b0 MacOS file system capitals got me. Will rename filter to Filter later Quadro 2021-05-18 16:36:58 -04:00
  • 84e246a963 Allow dimensions to be periodic Quadro 2021-05-18 16:36:29 -04:00
  • c18025c0b8 Luscher's filter - will tune this later Quadro 2021-05-18 16:35:39 -04:00
  • a918955020 Typo in assert Quadro 2021-05-18 16:35:01 -04:00
  • 86f08c6b9a Added a check that the initial EOFA action agrees with |eta|^2, thus checking the quality of the rational approximation in the heatbath Christopher Kelly 2021-05-18 13:57:44 -04:00
  • 9f0271039f Completed implementation of Meofa method of ExactOneFlavourRatio pseudofermion action Added tests to tests/forces/Test_mobius_force_eofa.cc testing that the EOFA heatbath results in Phi = M^{-1/2} eta Christopher Kelly 2021-05-18 12:27:51 -04:00
  • e3c18ce872 Adding boundary det Peter Boyle 2021-05-15 09:06:58 -04:00
  • 07a61e8826 renamed Peter Boyle 2021-05-15 09:06:22 -04:00
  • 58cb7c0732 passes force test and conserves in HMC plaquette looks odd Peter Boyle 2021-05-15 09:05:34 -04:00
  • 10339fd775 Rename Peter Boyle 2021-05-15 08:53:39 -04:00
  • 9400c207f7 Force test passing for boundary det Peter Boyle 2021-05-15 08:38:26 -04:00
  • 1fa89a2e7d Numerator pseudofermion Peter Boyle 2021-05-15 08:37:47 -04:00
  • 09b233b82e RdagR inverse Peter Boyle 2021-05-14 22:21:23 -04:00
  • 34ca4dd53a Nexp=20 default. Need a test for this. Peter Boyle 2021-05-14 22:20:41 -04:00
  • c19cf46169 Cayley causes problems if the argument is exactly ZERO Peter Boyle 2021-05-14 22:19:22 -04:00
  • c8db9ddb33 Massively improved test Peter Boyle 2021-05-14 21:56:24 -04:00
  • 104986b53d Prep for storing refresh actoin Peter Boyle 2021-05-14 12:02:15 -04:00
  • 91fd44419b Prep for storing action from the pseudofermion refresh Peter Boyle 2021-05-14 11:52:11 -04:00
  • d1daa0e3f7 Allow DD only in subset of dimensions Peter Boyle 2021-05-14 11:47:27 -04:00
  • 05e1aed326 commenting Peter Boyle 2021-05-14 11:45:06 -04:00
  • d3fd23117a Better block dimension checking and if Block >= Global dims implement non-Dirichlet Peter Boyle 2021-05-14 11:28:24 -04:00
  • def51267e9 Starting the DDHMC filter. This will be seriously subject to tuning as force in Omega and force in Omega bar appears to have different structure Peter Boyle 2021-05-14 11:26:47 -04:00
  • 5b52f29b2f Prepare to store the action in refresh Peter Boyle 2021-05-14 11:18:49 -04:00
  • 25bd03f201 Add LorentzVector Peter Boyle 2021-05-14 10:48:45 -04:00
  • d5edd100a5 Useful debug fingerprints Peter Boyle 2021-05-14 10:48:14 -04:00
  • 24df770f74 Added tests/IO/Test_field_array_io.cc testing/demonstrating parallel IO of an array of 5D fermion fields Christopher Kelly 2021-05-13 12:32:45 -04:00
  • 45b6c7effc Added a test code forces/Test_gpdwf_force_1f_2f that compares the action and force for DWF, EOFA and DSDR actions between the 1f and 2f implementations of G-parity BCs Broke up ExactOneFlavourRatio refresh into a virtual routine that generates eta and one that uses it as with the ratio and RHMC actions Added accessors to the pseudofermion field to TwoFlavourEvenOddRatio and ExactOneFlavourRatio Christopher Kelly 2021-05-12 16:34:07 -04:00
  • e39e326b79 sRNG compile fix Peter Boyle 2021-05-12 11:00:43 -04:00
  • 8458e13a23 Improved - untested Peter Boyle 2021-05-12 11:00:04 -04:00
  • 3575278b57 Small change Peter Boyle 2021-05-12 10:57:04 -04:00
  • 69a2c8769a Testing out the DDHMC constituent parts Peter Boyle 2021-05-11 13:56:52 -04:00
  • d4eaea00cf 2flavour local determinant ratio Peter Boyle 2021-05-11 13:56:21 -04:00
  • 347ccdc468 Delete file, replaced by SchurFactoredFermionOperator Peter Boyle 2021-05-11 13:55:52 -04:00
  • bf034ce239 The Dirichlet BC filter Peter Boyle 2021-05-11 13:55:27 -04:00
  • 791d0ab0b5 Provides the DDHMC lexicon for a pair of non-Dirichlet and Dirichlet operators Peter Boyle 2021-05-11 13:54:49 -04:00
  • 94a2a645bd Dirichlet BC wrapper for any Fermion Operator Peter Boyle 2021-05-11 13:54:15 -04:00
  • 281b55df04 srng changes Peter Boyle 2021-05-11 13:52:24 -04:00
  • a36e797bfc Use Integer Peter Boyle 2021-05-11 13:52:05 -04:00
  • 0bade717bf Added the momentum filter to resources Peter Boyle 2021-05-11 13:50:38 -04:00
  • 84fe791519 put momentum filter into the HMC resources Peter Boyle 2021-05-11 13:49:49 -04:00
  • e009a37f6e Moving the code to SchurFactorizedFermionOperator Peter Boyle 2021-05-11 13:46:40 -04:00
  • d7a887baf1 Better include guard Peter Boyle 2021-05-11 13:45:53 -04:00
  • 060bb59535 GCC warnings Peter Boyle 2021-05-11 13:38:27 -04:00
  • bac36399c1 Compiles Peter Boyle 2021-05-07 00:36:28 +02:00
  • b5b930d5bb Merge branch 'feature/ddhmc' of https://github.com/paboyle/Grid into feature/ddhmc Peter Boyle 2021-05-06 23:51:25 +02:00
  • 4fca66a7c6 Implement Dirichlet Comms option Peter Boyle 2021-05-06 23:44:45 +02:00
  • bd181b9481 Merge branch 'feature/ddhmc' of https://github.com/paboyle/Grid into feature/ddhmc Peter Boyle 2021-05-06 23:42:39 +02:00
  • 919ced1c31 Move the momentum filter to a better location for DDHMC Peter Boyle 2021-05-06 23:29:03 +02:00
  • b32fd473f8 Start at the Domain decomposed supprt Peter Boyle 2021-05-06 23:28:28 +02:00
  • ffcab64890 DDHMC filter Peter Boyle 2021-05-06 23:27:57 +02:00
  • 374fb325f3 Remove LIME dependency Peter Boyle 2021-05-06 23:26:42 +02:00