Remy Prechelt (109f04b4) at 24 May 19:36
@pranav Do you still get this error on the master
branch? Can you confirm that it's still happening?
If not, we should close this issue and open it again if the error pops up in the future.
@Fan Are you still up for producing a well-documented helper function for simple constant density media like water? (or ice? salt?).
If you are, feel free to assign yourself! Otherwise I will close this issue since it's not really "actionable" (although I do agree with you that the Mixin design pattern can be confusing for new users and we do need some simpler constructors!)
@JAlameddine Do you still want to merge this MR?
Or do you want to wait until we have completed the major surgery on Cascade
discussed in the last meeting?
Remy Prechelt (cad89b3a) at 27 Feb 08:56
Remy Prechelt (3a7011db) at 27 Feb 08:56
Merge branch 'rprechelt-python-energyloss' into 'master'
... and 1 more commit
This closes #486 by adding support for loading energy loss processes from the Python library.
This closes #486 by adding support for loading energy loss processes from the Python library.
My problem is that when I load the corsika
module in python3 and use the Library method I get the following error.
tamember@tale001:~/software/CORSIKA8/corsika-work/examples$ python3
Python 3.6.9 (default, Jan 26 2021, 15:33:00)
[GCC 8.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import corsika
>>> lib = corsika.Library("vertical_EAS_outputs")
Unable to instantiate an instance of 'EnergyLoss' for a process called 'energyloss'
module 'corsika.io.outputs' has no attribute 'EnergyLoss'
The only way that I can get the error to go away is to move the energyloss
directory out of the vertical_EAS_outputs
parent directory.
Has this not been updated yet? Or is this working as intended?
Hi @SirPottsalot!
Once !424 (merged) has been merged (should be in the next few minutes), this will be fixed on the master
branch. You will have to update your clone and reinstall the Python library.
This closes #486 by adding support for loading energy loss processes from the Python library.
Remy Prechelt (cad89b3a) at 27 Feb 08:10
Add support for loading the energy loss processes.
Hi @SirPottsalot!
Max beat me to responding by about 5 seconds. This script is currently outdated. Clearly the documentation is also wrong for recommending this script.
I've added this to our work list.
Hi @SirPottsalot!
This should not be what is happening! C8 has been undergoing a major internal reorganization the last few weeks and this bug might have slipped through.
I will update this issue once I've taken a look at what is going on.
Remy Prechelt (220b9600) at 20 Dec 05:18
The current tracking algorithm, a basic leapfrog, can sometimes exceed c
and isn't a proper relativistic tracker. I think the next step is to move to the Boris integrator, described here, which is relativistic, symplectic, phase-space-preserving, and supports both electric and magnetic fields.
@Nikos Can you update this MR with clang-format
? The format tests are currently failing.
Once the latest tests pass, I'll review this again tomorrow.
The new 5layer_atmosphere
, take a fixed magnetic field vector as far as I'm aware. For large scale air shower simulations (i.e. stratospheric or orbital detectors like ANITA/POEMMA), the showers can be so long that they experience a magnetic field vector that changes significantly over the shower length.
I'd like to change the interface so that you give a callable to 5layer_atmosphere
that takes the particle position and returns the magnetic field. For constant simulations, I'm guessing the compiler will just inline it via RVO but we can check to make sure.