Conrad Chan
Conrad Chan
**Work in progess** Type of PR: modification to existing code Description: Profiling of MPI runs has shown that the OMP critical sections for sending cells over MPI is a bottleneck....
#258 reduced some of the duplicated output lines due to MPI. #268 clobbered some of these improvements so the output is noisy again. This is a real nuisance when running...
Currently, when the particle array allocation is exhausted, the code halts. #214 solves this problem by increasing the fixed allocation size, but this is in general an inefficient use of...
The default disc setup with `IND_TIMESTEPS=no` gives an RMS error of roughly `1.e-14` when comparing MPI and non-MPI runs. With `IND_TIMESTEPS=yes`, the error is `1.e-6`, suggesting that there is a...
To reproduce this issue: ``` $PHANTOM_DIR/scripts/writemake.sh shock > Makefile make; make setup; make diffdumps ``` Setup with default options: ``` ./phantomsetup shock ``` Reduce resolution for quicker testing by changing...
Adding a build test for `HDF5=yes` would prevent issues like #137.
There are currently no version checks to ensure compatibility between Phantom and @cpinte's MCFOST. The documentation says to compile libmcfost.a and build Phantom using it, but there is no guarantee...
#128 implements periodic pings, but outputs an info-level log which is excessive and clutters the logs. As suggested by https://github.com/kpsuperplane/homebridge-wiz-lan/pull/128#issuecomment-1381898222 and https://github.com/kpsuperplane/homebridge-wiz-lan/issues/127#issuecomment-1382258065, these log messages should be for debugging only....
We're interested in using PySlurm to gather realtime statistics on jobs, using the `TRESUsageInTot` and `TRESUsageInAve` fields. These are available in the current database API (`pyslurm.db.JobStatistics`) but those are only...