Closed cdeil closed 8 years ago
I don't think by releasing 0.4 we claim that everything is nicely documented. IMO we should just get this out and worry about making everything nice and clean for future releases (latest: 1.0). Why should users start coming just because we put a 0.4 flag on the software?
@joleroi - I do think a few hours of effort to get our docs in order before cutting the 0.4 release is important. The most important thing is to have a few simple working examples that people can execute and play with to get started.
I just now fixed gammapy-image-crab.ipynb (data store changed, and now uses SkyMap).
@adonath - FYI: the SkyMap.plot
method doesn't work (when used without options) and .fill
is really awkward because it can't be called repeatedly to just stack counts.
@joleroi - spectral_analysis.ipynb is broken. And I would suggest you change to
from gammapy.data import DataStore
data_store = DataStore.from_dir('$GAMMAPY_EXTRA/datasets/hess-crab4-hd-hap-prod2/')
to access the data ... I think that has a better chance of new users setting it up correctly and understanding how it works.
@adonath - the ts_map_computation.ipynb example is broken. It's probably very simple updates for you. Can you please fix?
the spectral analysis notebook is fixed
@adonath @joleroi - Thanks for fixing the example notebooks.
I've improved the getting started section (with feedback from @adonath today) in #519 , see here.
The main improvement is that it now points users to the Gammapy notebooks and those are getting better.
Will make the release now ...
Gammapy 0.4 conda package is now available:
I'm not going to send out announcements for this ... 0.5 coming very soon.
@joleroi @adonath @leajouvin - I plan to cut the Gammapy 0.4 release tomorrow. We can use this issue to discuss it.
Current list of remaining issues and pull requests for the 0.4 milestone: https://github.com/gammapy/gammapy/milestones/0.4
With this release we basically have classical analysis, i.e. making and fitting 2d images and 1d spectra. But it's not documented nicely, so still very hard to use and get started for users.
I'll try to find time to work on the docs, but if you have time to make a pull request for some high-level analysis documentation, please leave a comment here and do.
Also -- please leave a comment here or file separate issues if there's something you think should be done for 0.4.