-
Notifications
You must be signed in to change notification settings - Fork 26
Planning Meeting 2020 03 18
Laura Hayes edited this page Apr 2, 2020
·
9 revisions
Attendance: @samaloney @hayesla @DanRyanIrish @rschwartz70 @binchensun @ianan @KriSun95
Topics
- HDEE proposal (Are the due dates changing?)
- PR statuses - (pull request!!!)
- Goals about compatibility, equivalence to current IDL
- Concrete test cases for RHESSI, NuStar, Fermi ...? --Adding a specific response matrix should be feasible and not difficult. So what we mean by a model would have parameters input for the electron function and the output should be counts in channels
- Repo stuff actions, tests, pipelines
- Strategic Package Organization
- Strategic Infrastructure Usage
- HDEE proposal discussion - lets go ahead with it.
- How are we going to organize the package - still at a stage to get to something thats works. Trying to develop how we ultimately want to look. Want it to play nicely with other python infrastructure. Learn as we go.
- Breaking out, or making it more modular the design of the codes (e.g. f_vth).
- What is the rundown in astropy - model composition part - does not include DRM - need to build that into the model that you want to fit.
- If you inherent from a base class - then changing this down the road is very difficult - lets be nimble.
- Fundamentally still have issues with response matrices and the model.
- Photon space model but also need this extra instrumental part.
- Instrument agnostic - then use small packages for each instrument with response matrix in the standard X-ray format.
- Breadboarded all the way through NuStar and astropy fitting and get a demonstration - which would be very powerful for the HDEE -
- Two NuStar events - one thermal (Kris) and also (non-thermal too) Lindsay that we can play around with.
- Create issue about everything to keep track of everything - don’t need a random PR
- Do we want to get the SSWIDL-->python codes then file them away and start again in a more ‘pythonic’ format.
- Shane’s PR is working - more tests to be written and then will be reviewed and merged
- Find some suitable RHESSI or new star data - some kind of python structure along with its response matrix- where should they go?
- Host data
- Shane’s PR
- Stay communicating - riot/email.
- Week after (this time the following week)
- Get repo set up with infrastructure types of things - CI etc.
- Example of Kris' data https://github.com/KriSun95/nustarFittingExample/blob/master/nustarFittingExample/NuSTAR%20Spectrum.ipynb