Comments (18)
Hi @MatthieuSchaller , I'll look into this later tonight.
from velociraptor-stf.
Actually, it looks like only the first 8 calculation options are actually implemented. There is also some strange overflow + modulo combinations in substructureproperties.cxx when setting the functions in lines 6335 and following.
from velociraptor-stf.
To be more specific, here is an example:
- Input file: /cosma7/data/dp004/jlvc76/COLIBRE/cold_gas/colibre_0023.hdf5
- Config file: /cosma7/data/dp004/jlvc76/COLIBRE/cold_gas/vrconfig_3dfof_subhalos_SO_hydro.cfg
- Version: 5eb6351
- Compilation line:
cmake ../ -DVR_USE_HYDRO=ON -DCMAKE_BUILD_TYPE=Debug -DVR_OPENMP=OFF
- Runtime command:
stf -I 2 -i colibre_0023 -o haloes -C vrconfig_3dfof_subhalos_SO_hydro.cfg
To get a version that works, comment out line 179 of the configuration and comment in line 180.
from velociraptor-stf.
One additional related thing. If you add aperture_total
to a list of things that contain other operations then the code does not crash but returns an array of 0s in the catalogue.
For instance
BH_internal_property_names=SubgridMasses,SubgridMasses,SubgridMasses,SubgridMasses,
BH_internal_property_input_output_unit_conversion_factors=1.0e10,1.0e10,1.0e10,1.0e10,
BH_internal_property_calculation_type=max,min,average,aperture_total,
BH_internal_property_output_units=solar_mass,solar_mass,solar_mass,solar_mass,
does not crash but produces and array of 0 for the last quantity whilst the first three are fine.
from velociraptor-stf.
Hi @rtobar @MatthieuSchaller
I think I've fixed all the bugs for apertures. The seg fault when running with only apertures loaded was a silly bug that is fixed. I've tested the branch with colibre data, loading in densities for -DVR_MPI=OFF
and mpi on. Seems to work. I will do some further testing but the branch is now available. @rtobar you could cherry pick the commits to a local ICRAR branch.
https://github.com/pelahi/VELOCIraptor-STF/tree/bugfix/load_extra_propeties_for_apertures
from velociraptor-stf.
Nice. Thanks very much for the prompt response!
I am happy to test it as well. Note that beyond the segfault (which was more of a symptom I think) the properties arrays ended up filled with zeros.
from velociraptor-stf.
The zeros should also be fixed. Based on the colibre snapshot, the values also seemed reasonable too.
from velociraptor-stf.
I'll try to cherry-pick the changes into a new issue-72
branch (gave it a quick crack, but needs some manual merging, so I'll have to look a bit closer into it). After that we can test the changes in cosma against the original inputs.
from velociraptor-stf.
@MatthieuSchaller: @pelahi's changes have been pushed now to the issue-72
branch, please try them out at your earliest convenience.
from velociraptor-stf.
Just tried that branch on the example above. It dies with:
[0000] [ 1.668] [ info] hdfio.cxx:2447 Reading file 0
[0000] [ 1.668] [debug] hdfio.cxx:1244 Opening Dataset PartType0/Coordinates
[0000] [ 1.669] [debug] hdfio.cxx:1244 Opening Dataset PartType1/Coordinates
[0000] [ 1.669] [debug] hdfio.cxx:1244 Opening Dataset PartType4/Coordinates
[0000] [ 1.669] [debug] hdfio.cxx:1244 Opening Dataset PartType5/Coordinates
[0000] [ 1.669] [debug] hdfio.cxx:1244 Opening Dataset PartType0/Velocities
[0000] [ 1.669] [debug] hdfio.cxx:1244 Opening Dataset PartType1/Velocities
[0000] [ 1.669] [debug] hdfio.cxx:1244 Opening Dataset PartType4/Velocities
[0000] [ 1.670] [debug] hdfio.cxx:1244 Opening Dataset PartType5/Velocities
[0000] [ 1.670] [debug] hdfio.cxx:1244 Opening Dataset PartType0/ParticleIDs
[0000] [ 1.670] [debug] hdfio.cxx:1244 Opening Dataset PartType1/ParticleIDs
[0000] [ 1.670] [debug] hdfio.cxx:1244 Opening Dataset PartType4/ParticleIDs
[0000] [ 1.670] [debug] hdfio.cxx:1244 Opening Dataset PartType5/ParticleIDs
[0000] [ 1.670] [debug] hdfio.cxx:1244 Opening Dataset PartType0/Masses
[0000] [ 1.670] [debug] hdfio.cxx:1244 Opening Dataset PartType1/Masses
[0000] [ 1.670] [debug] hdfio.cxx:1244 Opening Dataset PartType4/Masses
[0000] [ 1.671] [debug] hdfio.cxx:1244 Opening Dataset PartType5/DynamicalMasses
[0000] [ 1.671] [debug] hdfio.cxx:1244 Opening Dataset PartType0/InternalEnergies
[0000] [ 1.671] [debug] hdfio.cxx:1244 Opening Dataset PartType0/StarFormationRates
[0000] [ 1.671] [debug] hdfio.cxx:1244 Opening Dataset PartType0/MetalMassFractions
[0000] [ 1.671] [debug] hdfio.cxx:1244 Opening Dataset PartType4/MetalMassFractions
[0000] [ 1.671] [debug] hdfio.cxx:1244 Opening Dataset PartType4/BirthScaleFactors
[0000] [ 1.672] [debug] hdfio.cxx:1244 Opening Dataset PartType4/BirthDensities
[0000] [ 1.672] [debug] hdfio.cxx:1244 Opening Dataset PartType5/SubgridMasses
HDF5-DIAG: Error detected in HDF5 (1.10.6) MPI-process 0:
#000: H5Dio.c line 185 in H5Dread(): could not get a validated dataspace from file_space_id
major: Invalid arguments to routine
minor: Bad value
#001: H5S.c line 255 in H5S_get_validated_dataspace(): selection + offset not within extent
major: Dataspace
minor: Out of range
Error in HDF routine
Abort(1) on node 0 (rank 0 in comm 0): application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
Note in particular that the new fields I am asking to read are not listed in Opening Dataset xxx
. However, if I revert to using another kind of property calculation (say average) then it is listed there and the code runs.
from velociraptor-stf.
I think I've fixed the MPI error. There's an update to the branch
from velociraptor-stf.
@MatthieuSchaller I've just cherry-picked the latest update from @pelahi and put it in the issue-72
branch for you to test.
from velociraptor-stf.
Thanks!
The code does not crash anymore at that point. However, the fields are still not listed so I am not sure they are actually read in.
FYI, I am compiling with -DVR_USE_HYDRO=ON -DVR_OPENMP=OFF -DVR_MPI=ON
.
from velociraptor-stf.
The code completed succesfully and the arrays are filled with non-zero values. I'll check them today.
from velociraptor-stf.
All good! The new properties are all computed correctly.
from velociraptor-stf.
As an additional check, this also now worked correctly when using OpenMP.
from velociraptor-stf.
Thanks a lot @pelahi for providing the fixes for this problem, and to @MatthieuSchaller for initially reporting and now verifying the fixes. After a final test round via #74 I merged the issue-72
branch into the master
, so we can probably declare this issue as fixed.
from velociraptor-stf.
As always, thanks for the prompt and efficient response guys! Much appreciated.
from velociraptor-stf.
Related Issues (20)
- Incorrectly sized buffer given for MPI_Bcast reception HOT 1
- Writing parallel properties file in hydro builds is broken HOT 6
- SO list offsets are wrong/counterintuitive HOT 6
- Inconsistent array names between properties files HOT 6
- SO list output too large and possibly wrong HOT 11
- Error in writing HDF5 outputs HOT 5
- Improve VR's memory usage for extra data in Particles
- DMO Zoom on-the-fly with SWIFT segfault HOT 4
- OpenMP bug in temperature calculations. HOT 5
- Memory usage blowing up in large DMO runs HOT 22
- Differences in halo masses when switching on/off substructure search HOT 2
- Mistakes in metallity calculations HOT 1
- Apparently wrong output when using Star_internal_property options
- HIGHRES needs undocumented Extensive_interloper_properties_output config option
- Error on compiling HOT 9
- Particle_type_for_reference_frames value not listed in the output HOT 3
- Crash with large SWIFT runs using MPI HOT 2
- Very wrong results on a zoom.
- Crash in MPIBuildParticleNNExportListUsingMesh() HOT 2
- Buffer overflow in PotentialTree with OpenMP HOT 3
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from velociraptor-stf.