aboutsummaryrefslogtreecommitdiff
path: root/Carpet/CarpetIOHDF5
Commit message (Collapse)AuthorAge
* CarpetIOHDF5: avoid variable shadowing.Anton Khirnov2013-11-05
|
* CarpetIOHDF5: remove a superfluous H5Pclose call.Anton Khirnov2013-11-05
|
* CarpetIOHDF5: correct output method name in INFO messageRoland Haas2013-09-25
|
* CarpetIOHDF5: check whether map exists before accessing itRoland Haas2013-08-18
| | | | | | | | | | this allows data files from multipatch runs to be read in with the file reader into cartesian runs if only the inner cartesian patch is required.CarpetIOHDF5: check wehter map exists before accessing it this allows data files from multipatch runs to be read in with the file reader into cartesian runs if only the inner cartesian patch is required.
* CarpetIOHDF5: remove two compiler warningsRoland Haas2013-08-18
|
* CarpetIOHDF5: fix reading of new-style grid structuresRoland Haas2013-08-18
| | | | | | | | last commit to CarpetIOHDF5 broke this. This commit also updates the test suite data so that it actually tests the file format. This commit adds a level 2 warning if no grid structure is found in file.
* CarpetIOHDF5: use simple_extents rather than storage size on diskRoland Haas2013-08-12
| | | | | | | | | | storage size is literally the amount of space used on disk, so if eg compression is used, this is much smaller than the amount of space required to hold the data in memory. Also change type of data read in to what the memory dataspace is not what the dataspace on disk is. This way HDF5 actually converts from the on-disk representation to the in-memory one.
* CarpetIOHDF5: support IO->alias option in readerRoland Haas2013-08-10
| | | | | this allows the reader to read a dataset into a different variable than which one was written. Eg. GRHydro::dens in PPAnalysis::dens.
* CarpetIOHDF5: delay browsing of input filesRoland Haas2013-08-10
| | | | this means the BrowseDataSets is only called from one location
* Timers: Move all timer-related code into a new thorn TimersErik Schnetter2013-08-07
|
* CarpetIO*: New parameter CarpetIO*::out_group_separatorErik Schnetter2013-05-26
| | | | out_group_separator chooses the string by which thorn name and group name are separated in file names. The default is "::" for backward compatibility. This parameter only affects output where CarpetIO*::one_file_per_group is set; otherwise, the thorn name does not appear in the file name.
* Replace Cactus complex number type with C/C++ complex numbersErik Schnetter2013-03-08
| | | | | | | | | | | | | Map CCTK_COMPLEX to "double complex" in C, and "complex<double>" in C++. (It is already mapped to "double complex" in Fortran.) Update type definitions. Re-implement Cactus complex number math functions by calling the respective C functions. Update thorn that access real and imaginary parts of complex numbers to use standard-conforming methods instead.
* CarpetIOHDF5: remove index files along with checkpoint filesRoland Haas2013-02-28
|
* CarpetIOHDF5: Make variable unsigned to avoid compiler warningErik Schnetter2013-01-18
|
* CarpetIOHDF5: Remove unused variables, initialise other variable to avoid ↵Erik Schnetter2013-01-18
| | | | compiler warnings
* CarpetIOHDF5: Sliced HDF5 output index file not stored correctlyDavid Radice2012-12-19
| | | | | | | | Index HDF5 datafiles weree not handled correctly in the case of sliced data. The files are created and initialized correctly at the first iteration, but every subsequent access fails with an HDF5 error. Patch by David Radice.
* CarpetIOHDF5: comment unused variable (commented because code using it is ↵Frank Löffler2012-11-29
| | | | also commented
* CarpetIOHDF5: cast string constant to char array to get rid of compiler warningFrank Löffler2012-11-29
|
* Allow padding in transport operatorsErik Schnetter2012-11-22
| | | | | Rewrite padding infrastructure. Add padded array extents to transport operator APIs.
* CarpetIOHDF5: fix check for missing datasets when called in FileReader modeRoland Haas2012-10-29
| | | | only check datasets whose variables were actually requested to be read
* CarpetIOHDF5: close all HDF5 objects when output file is closedRoland Haas2012-10-24
| | | | also garbage collect HDF5 at each H5close
* CarpetIOHDF5: output index files for sliced output if requestedRoland Haas2012-10-20
|
* CarpetIOHDF5: make index files sparse files with full datasetsRoland Haas2012-09-11
| | | | | | | | | rather than datasets of extent [1,1,1]. It so happens that this patch removes the separate shape arrays and separate dataspaces for the datasets in index files and creates the index file datasets with the same dataspace as the "heavy" file datasets. It retains the h5shape attribute that was originally introduced for index files even though it is now redundant (since one could call H5Sget_simple_extent_dims on the datasets in the index files).
* CarpetIOHDF5: Use string instead of char[] in HDF5 for large attributesErik Schnetter2012-09-11
|
* CarpetIO*: Do not overwrite last_output_iteration after recoveryErik Schnetter2012-09-11
|
* CarpetIOHDF5: Increase array size for last_output_iteration_sliceErik Schnetter2012-09-11
|
* CarpetIOHDF5: close all HDF5 objects when closing a fileRoland Haas2012-09-11
| | | | garbage collect HDF5 objects when closing files
* CarpetIOHDF5: add support for index files to file readerRoland Haas2012-09-11
| | | | | | we try for index files when opening the first file and if this does not succeed do not try opening index files again. This reduces the number of file system accesses when no index files are present.
* CarpetIOHDF5: teach the checkpointer to write index filesRoland Haas2012-09-11
|
* CarpetIOHDF5: defer opening data files until we have to to read aRoland Haas2012-09-11
| | | | dataset
* CarpetIOHDF5: support indices in sequential chunked outputRoland Haas2012-09-11
| | | | which happens to be the output method used for single process runs.
* CarpetIOHDF5: do not re-parse files for patches when usingRoland Haas2012-09-11
| | | | open_one_input_file_at_a_time
* CarpetIOHDF5: close memory and HDF5 properties leaks mostly associatedRoland Haas2012-09-11
| | | | with open_one_input_file_at_a_time
* CarpetIOHDF5: add option checkpoint_every_divisorRoland Haas2012-09-11
| | | | | | to checkpoint when cctk_iteration % checkpoint_every_divisor == 0 rather than at whenever checkpoint_every iterations have passed since the last checkpoint
* CarpetIOHDF5: only warn once about reading extra filesRoland Haas2012-09-11
| | | | | once per variable, refinement level and time level that is. So still about ~3000 warnings in a typical simulation.
* Move MPI support from flesh to a thornErik Schnetter2012-09-11
| | | | | | | Move MPI support from flesh to thorn ExternalLibraries/MPI. This also requires thorns that call MPI directly to declare this in their configuration.ccl. Existing configurations using MPI need to include ExternalLibraries/MPI into their thorn list.
* CarpetIOHDF5: add test for buffer zone outputRoland Haas2012-09-11
| | | | Right now there is no facility to actually use this test unfortunately.
* CarpetIOHDF5: re-add ghost and symmetry points when removing buffersRoland Haas2012-09-11
| | | | | | | NOTE: this assumes (like other parts of CarpetIOHDF5) that the number of symmetry points is the number of ghost points. NOTE: it likely outputs too many points when RotatingSymmetry is used and only buffer points touch the symmetry boundary.
* CarpetIOHDF5: zero cctk_nghostzones attribute unless ghost zones are writtenRoland Haas2012-09-11
| | | | to disk
* Remove all #warning statementsErik Schnetter2012-09-11
|
* CarpetIOHDF5: use version 1.6 of the HDF5 API within hdf5_recombinerknarf2012-09-11
| | | | | | | There are still systems with only version 1.6 installed, or broken installs of version 1.8 (Debian system packages at least up to squeeze). These systems benefit from a compiling hdf5_recombiner, while version 1.8 is not really required to be used here.
* CarpetIOHDF5: Simplify makefileErik Schnetter2012-09-11
|
* CarpetIOHDF5: More internal error checkingErik Schnetter2012-09-11
|
* CarpetLib: Change API to obtain pointer to grid function dataErik Schnetter2012-09-11
| | | | | | | | Change the API to obtain a pointer to grid function data: - Use a function "typed_data_pointer" instead of overloading the () operator (because this looks nicer) - Don't use a virtual function (because this isn't needed) - Update all uses
* CarpetIOHDF5: hdf5toascii_slicer take patch.iorigin into account when ↵Roland Haas2012-09-11
| | | | | | | | | | | computing coordinates of points this is the same issue (just seems from the other side of the output) as in "CarpetIOHDF5: Correct iorigin attribute for 2D output", namely that iorigin is stored in multiples of the stride for the given refinement level --- Carpet/CarpetIOHDF5/src/util/hdf5toascii_slicer.cc | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-)
* hdf5_slicer: close input files as soon as possible to conserve memory and ↵Roland Haas2012-09-11
| | | | | | | | file descriptors --- Carpet/CarpetIOHDF5/src/util/hdf5_slicer.cc | 13 +++++-------- 1 files changed, 5 insertions(+), 8 deletions(-)
* CarpetIOHDF5: let the hdf5_slicer write out a cctk_nghostzones attribute whichroland haas2012-09-11
| | | | is used by VisIt
* mergeknarf2012-09-11
|\
| * CarpetIOHDF5: Update to new gdata::copy_from APIErik Schnetter2012-09-11
| |
| * CarpetIOHDF5: Introduce API to checkpoint only a subset of groupsErik Schnetter2011-12-14
| | | | | | | | | | | | | | Introduce a new API to checkpoint only a subset of group, via an aliased function IO_SetCheckpointGroups. This can be used for simulation spawning, i.e. off-loading certain calculations (e.g. analysis) outside of the main simulation.
generated by cgit v1.2.3 (git 2.39.1) at 2024-05-17 02:17:48 +0000