Christoph Deil

> I found the point raised by @giacomov on speed quite relevant: would there be a way to improve the sherpa modeling package for speed in a way acceptable for...

I don't have time to look into Sherpa before our call on Friday. So for now I've just asked here: https://github.com/sherpa/sherpa/issues/184

Some updates: - There's an open pull request to add quantity support to Astropy modeling. - There's a proposed GSoC 2016 project to "Bridge sherpa and astropy fitting" http://openastronomy.org/gsoc/gsoc2016/ideas.html#Astropy -...

There's discussion here at [PyAstro16](http://python-in-astronomy.github.io/2016/) how to move forward with modeling for astronomers. We're now starting to collect notes and use cases (preferably, but not necessarily with working code) here:...

Well, in HESS, 100 people have done all kinds of studies for over a decade, so I'm sure many studies and attempts to quantify IRF and high-level analysis result precision...

Also -- every IACT, chain, config and IRF is different. See the backup slides 12 and 13 with examples of IRF issues from HESS [here](https://github.com/open-gamma-ray-astro/2016-04_IACT_DL3_Meeting/blob/master/presentations/2016-04_IACT_DL3_Meeting_Deil_HESS.pdf). Some chains in HESS do...

> But, thinking about CTA, the fact that we need to validate the format itself worries me. I guess it may be validated together with the methods that will, in...

I don't have a strong opinion either way. Yes, using `y=log(e_reco)` is a bit simpler and avoids e.g. the clipping issue. But the clipping can also be avoided by just...

@mdpunch - Thanks for the comment about the `yy = e_true / e_reco < 1` proposal and mentioning it's advantage. One comment concerning RMF: it has this kinda complex zero-suppression...

> Note that in any case, we should also store some sort of error bars with the matrix (something we learned when doing sensitivity curves). The reason for that is...