Add recipe for simulating synthetic lightcurves and fitting the PSD of an observed curve.

[cgalelli]

This PR adds a recipe which implements some functions to generate synthetic lightcurves, nominally:

• the TK95 algorithm that will be added in the main gammapy code with 1.3
• the more complex EMM algorithm

It also adds the function to create envelopes of the periodogram for a lightcurve, and a fitting function for the PSD of an observed curve. These function can also be found in the small package gammapy_SyLC that will probably be adapted to be used here in the future, once problems with installation are resolved.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[bkhelifi]

Thanks a lot @cgalelli ...
@registerrier, should we precise in some headers the author, the date and the used Gammapy version?

[registerrier]

Thanks @cgalelli . This looks good to me. It would be simpler if some of the code could be imported for SyLC indeed. We can keep that for later though.

[registerrier]

Thanks a lot @cgalelli ... @registerrier, should we precise in some headers the author, the date and the used Gammapy version?

The gammapy version is defined in the env.yml. Maybe a print(gammapy.version) would be sufficient.

Regarding names, we have not done it yet, but we could/should. What would be the proposal for that? How does additional changes (maintenance, upgrades) is taken into account?

[cgalelli]

I finally could solve the problem with the import of gammapy_SyLC so the new version has everything implemented as a package function. Changed the text accordingly. Also added the author line at the end.

[registerrier]

Thanks @cgalelli . There seems to be a problem with the notebook. It is not recognized as such in the CI. Could you check if there is a problem? Is the very long name a good idea?

[cgalelli]

Thanks @registerrier , indeed i had made a bit of a mess in the restructuring of the notebook. Now it should be fine. I had also used an old local name for the notebook that was quite long and cumbersome - returned to the name I had in the first commit.

The CI I think will still fail due to the import of gammapy_SyLC

[cgalelli]

Apparently it does not: i am trying to find an alternative way of installing the package in the environment file.