Loving ISIS - Confessions of a Former XSPEC User

Power Tools:

The examples here have just really scratched the surface of the capabilities of ISIS. As described in the ISIS manual, there's lots of very useful stuff that it can do with the APEC database. There's also all sorts of 'advanced methods' for setting up data and models. For example, one can assign multiple responses to the data (e.g., multiple gratings orders for HRC-LETG data). You can combine data sets (see the .help file for the ISIS command combine_datasets - a more general way of combining data than adding there spectra and averaging there responses outside of the analysis package). The latter can be used to model sources that are very close together on a detector image, i.e., where one expects their spectra to 'overlap', etc.

One of the more exciting prospects is the use of parallelization to run computationally expensive error bars and fit functions. Work is going on here to make that very easy and straightforward to set that up. Description of how you can achieve super-linear speed ups in your error bar searches can be found on this ISIS web page!

As others write web-pages describing these advanced ISIS features and capabilities, I will provide a list of links below.

• aglc - A suite of S-lang routines from Dave Huenemoerder to perform timing analysis on Chandra gratings data. Although not solely an ISIS set of tools (it can be run under Sherpa or under the slsh S-lang shell, for instance), it does make use of ISIS intrinsic functions, as well as the histogram module from the S-Lang Modules Packages.
  
  
• SITAR - A suite of S-lang routines to perform timing analysis, including the ability to define an averaged, binned Power Spectrum as an ISIS or Sherpa dataset, suitable for fitting. Again, not solely an ISIS set of tools, but it does rely upon ISIS intrinsic functions.
  
  
• triplet_fit.sl - Routines from Dave Huenemoerder to help model Helium-like triplets in Chandra gratings data.