Use micromamba instead of conda for faster builds - #81
Conversation
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
|
The notebook that is failing is just an ERDDAP server timeout. It works locally as expected. This one is good to go. |
|
@MathewBiddle the notebook 2016-11-15-glider_data_example.ipynb changed significantly. Iris no longer accept missing ancillary data, so most of our dataset will just crash with it. I took this opportunity to re-write the notebook using |
|
Looks like I'm running into an issue with cf_xarray in glider notebook.
|
Can you send me you the results of Also, can you try to append
before reading? That issue should've been solved but you either have an older version of netcdf-c there or the problem resurfaced. |
Adding |
|
BTW, you need to update your env. You have |
|
But, it's Friday! Not sure I want to open that can of worms. Wish me luck 🤞 |
|
after the update/reinstall the glider notebook is running smoothly again. Sorry for that. |
|
Looks like iris is barking in the CF-UGRID notebook too. |
Not in the CIs, right? Are you running it locally? If so, I'll need your |
|
Yes, I am testing it locally. Just to make sure folks would be able to run things locally. |
It works for me locally and in the CIs :-/ Maybe it is a Windows thing or just a temporary network problem. Did you try running it again? |
|
I did with no luck. I'll try a "turn it off and back on again" and see if that fixes it. |
|
Ugh, after reboot a new error: I'll work through this next week. |
That one is a known bug in udunits when the installation is in a different drive from the one you are running the notebook. See Unidata/UDUNITS-2#43 |
|
Interesting. That went away when I updated udunits2. Now I'm back to the error above #81 (comment) Odd that a fresh install would give me these issues. 🤔 |
🤣 The "new" udunits2 was never really released but I guess we patched it and forgot. It is still odd that you got the old one in a fresh env.
Can you give me more context? What cell does that happen? What is the full traceback? |
|
The first cell of the CF-UGRID-SGRID notebook returns this message: |
|
You have the same package versions I do and the same ones that are installed in the CIs :-/ I'll test it on Windows in a moment just to be sure. |
|
I can reproduce it. Quite puzzling b/c it does not happen on Linux with virtually the same environment. I'll debug this further but at this point I think I prefer to re-write this notebook and ditch iris altogether in the same fashion we did for the glide data notebook. |
|
Making it a draft until I release odvc. |
|
@MathewBiddle this one is ready for another round. You must install odvc >=1.0 first. |
|
Cool! After updating odvc everything is working again! Last request is to add a line |
Done. I'll also add those to the PR template so we don't forget to do that in the future. |
|
Looks good |
We should also cache the env too.