marx accuracy and testing

No simulation is ever perfect. In this section, we show tests for the internal consistency of marx and compare simulations to real observations. This demonstrates the limits to which marx simulations can be trusted. Anybody using marx for scientific analysis needs to consult the following pages to determine if the simulations are realistic enough for their purpose.

For example, the marx HRMA module uses a simplified geometry to speed up simulations so that the user can simulate millions of rays in less than a minute on a typical desktop computer. Thus, the simulated PSF will certainly differ from the observed PSF. SAOTrace has more details, but does not explain the “hook” feature of the PSF seen on very small scales either. Only a comparison between simulations and observed data can tell how well the simulations are doing.

• Point Spread Function (PSF)
  • On-axis PSF on an ACIS-BI chip
  • On-axis PSF on an ACIS-FI chip
  • On-axis PSF for an HRC-I observation
  • Off-axis PSF
  • On-axis PSF at different energies
  • Simulated off-axis PSF
• Wings of the Chandra PSF
  • Summary
• Flight grade distribution
  • Grades on an ACIS-BI chip
  • Grades on an ACIS-FI chip
• Reproducing an input spectrum
  • Absorbed powerlaw on ACIS-S
  • Powerlaw spectrum of an off-axis source
  • Two thermal components on ACIS-I, on-axis
• Source shapes in MARX
  • Build-in geometric sources
  • Image as source
  • Using a RAYFILE source
  • Compiling a USER source
• Coordinates on the sky and the chip
  • Chandra Orion Ultradeep project
  • Simulations with an IMAGE source
  • Regular Grid (ACIS)
  • Regular grid (HRC)

The code to run all these tests is available in the notebooks. Advanced users may wish to inspect the code of the tests for some more ideas on how to use marx. The code is lightly commented, but not in as much detail as Examples of MARX in use. To execute the notebooks, you will need to be in a CIAO environment with marx and SAOTrace installed. We also use astropy, a Python package that is not part of CIAO by default. You can install it using pip install astropy.