search

UofA-HEAG / CTA-Simulation

MIT License
2 stars 1 forks source link

Sim_telarry: Estimate computation time & disk storage #3

Open seinecke opened 5 years ago

seinecke commented 5 years ago

Let's do some test runs with different configurations to get some estimates on computation time and disk storage.

Process the different Corsika runs for the following configuration:

hdp1213 commented 5 years ago

I can't see a way of changing the telescope array geometry in sim_telarray, as I used the baseline 70 SST telescopes for the CORSIKA runs. From my understanding, CORSIKA (or rather, the IACT/ATMO add-on package) takes in the telescope array geometry and models them as a set of spheres and only saves the shower particles that intersect these, so sim_telarray only needs to use these particles in its simulation. It looks like I'll have to do some other CORSIKA runs with the same parameters but with the 5 SST layout and go from there.

hdp1213 commented 5 years ago

I've run sim_telarray on the new CORSIKA outputs from the 5 SSTs, and here's what I've been getting. With how quick the protons are done, I have my doubts things are working as intended, but then again I'm not sure what the ballpark for sim_telarray runtime is. The output files don't seem to have anything that jumps out at me as wrong, either. Worth noting is that all of these are done with 10 array/shower copies, as opposed to 20.

Configuration Computation time Disk storage
gamma, 5 SSTs, nshow=1000, max scatter radius = 1500m 06:03 38MB
protons, 5 SSTs, nshow=1000, max scatter radius = 1500m 00:17 1.5MB
protons, 5 SSTs, nshow=1000, max scatter radius = 1500m, viewcone=20 00:07 440KB