Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Skyfield VS Ephem #302

Closed
antvig opened this issue Dec 6, 2018 · 16 comments
Closed

Skyfield VS Ephem #302

antvig opened this issue Dec 6, 2018 · 16 comments

Comments

@antvig
Copy link

antvig commented Dec 6, 2018

I'm having trouble installing workalendar since it requires ephem which need to be compiled. (I can't use a precompiled conda package for my purpose.)

I've seen that there is the package skyfield which does the same than ephem and that doesn't require to be compiled.

Are you aware of skyfield? If yes, is there a reason why workalendar uses ephem instead of skyfield?

@antvig
Copy link
Author

antvig commented Dec 6, 2018

Okay, I just looked deeper at skyfield and it is based on Nasa files. So it needs to be downloaded first. I guess my above comment is not relevant.

@brunobord
Copy link
Member

I agree that somehow it's a bit painful to require development packages to use ephem.
I didn't know about Skyfield and it might be interesting to get into it. I'm not very enthusiast about having to download NumPy and other requirements, but if it removes the fact that you need Python-Dev somewhere, why not?

I'll keep your issue in my backlog, for later. thank your for the heads-up!

@antifa-ev
Copy link

@antvig: What's so difficult here? Just run python setup.py build and then python setup.py install --optimize=1

@smarie
Copy link

smarie commented Jan 9, 2019

On some targets (windows...) recompiling is much less friendly than on others ("let's install windows SDK ! and visual C++ redist ! oh no let's try mingw instead... oh wait the 64 bits version seems to have issues with cpython... oh wait I have to modify a file in the pip distrib to change some includes..." etc.)

So I understand that some users may just be afraid of compiling python packages on windows.

However @antvig can you explain why you can not use the precompiled "ephem" conda package in your case ? Thanks !

@brunobord
Copy link
Member

it looks like calculating solar terms using Skyfield is a bit harder than I first thought... stay tuned, I didn't say my last word!

brunobord added a commit that referenced this issue Jul 22, 2019
This Mixin was barely a mixin at all: none of its method was using ``self``.

it refers to #302, but not essentially. The goal of this refactoring is to simplify classes hierarchy, and it might also simplify the swap between `ephem` and `skyfield`.
brunobord added a commit that referenced this issue Aug 23, 2019
@brunobord
Copy link
Member

look, ma! no pyephem! :o)

brunobord added a commit that referenced this issue Aug 23, 2019
brunobord added a commit that referenced this issue Aug 23, 2019
@brunobord
Copy link
Member

I'm finally able to present a "PyEphem-free" version of workalendar.
I think this is an important requirement change, so I may need some volunteers to test a pre-version on their machine and tell me if it's okay.
/cc @antvig, @smarie (or even @antifa-ev) would you be interested?

brunobord added a commit that referenced this issue Aug 23, 2019
refs #302, #348

Many thanks to @GammaSagittarii for the tremendous help on finding the right way to compute Chinese Solar Terms.
@brunobord
Copy link
Member

In case any of you is interested in testing the next Workalendar Release Candidate, here's a procedure you may try to follow: #398

@brunobord brunobord added the blocked Blocked by a question, an undefined rule to compute a date, etc. label Aug 30, 2019
@smarie
Copy link

smarie commented Sep 1, 2019

Cool ! I let @antvig have a look if he can as he has been much more involved than I on this topic

@antvig
Copy link
Author

antvig commented Sep 3, 2019

I will certainly take some time to test it !

@brunobord
Copy link
Member

@antvig any news on this point?

@brunobord
Copy link
Member

I'm sorry to have to look a bit pushy, but I'd really want to have at least one more user to confirm that the release candidate is working.
I really wish to move forward on this long-due improvement and I depend on a Windows test (or two, or three, or well...).

So, who can confirm that the RC is working, following the procedure described in #398?

@antvig
Copy link
Author

antvig commented Sep 20, 2019

I just did at the moment ;)

brunobord added a commit that referenced this issue Sep 20, 2019
refs #302, #348

Many thanks to @GammaSagittarii for the tremendous help on finding the right way to compute Chinese Solar Terms.
@brunobord
Copy link
Member

WOOT!

@brunobord brunobord removed blocked Blocked by a question, an undefined rule to compute a date, etc. question labels Sep 20, 2019
brunobord added a commit that referenced this issue Sep 20, 2019
refs #302, #348
closes #398

Many thanks to @GammaSagittarii for the tremendous help on finding the right way to compute Chinese Solar Terms.
@brunobord
Copy link
Member

version 7.0.0 of workalendar has been released and published on PyPI. Upgrade away, to the stars, and beyond!

@smarie
Copy link

smarie commented Sep 23, 2019

Thanks @brunobord !

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants