-
Notifications
You must be signed in to change notification settings - Fork 123
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
[FTheoryTools] More on serialization of FTheoryTools #4008
base: master
Are you sure you want to change the base?
[FTheoryTools] More on serialization of FTheoryTools #4008
Conversation
b31d855
to
58ae7ad
Compare
10ed711
to
54b05b2
Compare
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #4008 +/- ##
==========================================
- Coverage 84.60% 84.58% -0.02%
==========================================
Files 631 631
Lines 84811 84830 +19
==========================================
+ Hits 71751 71755 +4
- Misses 13060 13075 +15
|
54b05b2
to
fafe0b4
Compare
82437b0
to
9299f46
Compare
@antonydellavecchia , as we discussed during the coding sprint in KL, I will move most of the above ToDos to #2699 and will then try to brush up this PR. |
3642acb
to
45d2216
Compare
@antonydellavecchia I have brushed up this PR. In particular, I have updated the QSMDB. I am not sure if we can modify the current QSMDB artifact (attached to a release tag of OSCAR) I have moved it to https://martinbies.github.io/Materials/QSMDB/qsmdb.tar.gz, where I can modify/edit it more conveniently. Not sure if this is the best move though and am happy to change this. Please let me know what you think. In any case - I would expect that with the current changes, all tests pass and thus, this PR is getting ready to be merged. |
45d2216
to
5e6e671
Compare
@antonydellavecchia I am puzzled why the tests fail. The new artifact files should not cause the problem in the failures. So I am wondering if somehow OSCAR fails to fetch those new artifact files? |
5e6e671
to
e5bfd89
Compare
@antonydellavecchia As discussed in slack, I have added the new files to the release tag now. Hopefully the tests do succeed now. |
e5bfd89
to
f2655f4
Compare
f2655f4
to
422aa16
Compare
Ok. I think I have finally succeeded. At least on my local system it works. I have renamed the previous QSMDB into QSMDB-old and added the new version as QSMDB. The latter is larger in size by about 1MB than the old file. I believe this is because the hypersurface models do remember more details now than they used to at the time when the initial database was created. |
I changed the name back (of the file in the github release), the old file must stay exactly as it is. It is still used in the existing oscar releases. |
The name of the artifact and the filename in the url do not need to match, with the files renamed again as mentioned above the following works fine:
Edit: That one download failure is probably because Pkg will try the package servers first and then use the specified url. |
Thank you @benlorenz ! I believe this PR is now ready to be reviewed. |
There are failures in the doctests |
Yeah, @benlorenz already commented on them above: "That one download failure is probably because Pkg will try the package servers first and then use the specified url." I do not know (yet) how to avoid this. Suggestions welcome. Hhhmm... just realized that there are other failures too... |
The failure I was referring to was just this line in my small example:
But this is resolved automatically by retrying:
|
56e1562
to
96da15a
Compare
The map creation (for I pushed a fix to this branch which appends another (identity) map to the correct class group in that case: |
Awesome! Very nice fix. Thank you very much indeed @benlorenz ! :) |
6604d6e
to
0de86dc
Compare
(Nitpick: I put the fix to class group as the first commit: In this way, I do expect that all commits work individually, and so the changes in this PR can be rebased onto the master branch, rather than squashed.) |
As mentioned in the other PRs, it seemed natural to group all the different serialization discussions we had in one PR. For this, I need your help @antonydellavecchia whenever you are ready (not urgent).
Points to be addressed in this PR include (but are not necessarily limited to the following):
The below points were initially meant to be fixed in this PR, but we (@antonydellavecchia and I) decided that it will be better to deal with them in separate PRs.
cc @antonydellavecchia @emikelsons @apturner