MRG: Build relevant examples #3365
Conversation
|
From the first commit, which only touched And from the second commit, which touched a couple of tutorials, we get the following run: Which produces: I like it, but I don't think we can merge this until the caching problem is solved. This build took an hour and 11 minutes, 57 minutes of which was downloading data (!). |
|
LGTM thx @Eric89GXL |
|
Okay I'm going to rebase, hopefully it picks up the cache and takes less time now. |
|
@jaeilepp looks like we still get: After this finishes successfully, I'll try pushing another commit. Maybe it needs a couple of tries to save properly. |
|
Still appears to have failed: I don't understand why it would get the other directories. That's weird. |
|
Yeah I'm trying to debug this by running minimal builds on my branch. Let's see if I can find something. |
|
I'm wondering if our cache is just stuck in some weird state and it would fix itself if we just clear it once. I can't think of anything else right now. |
|
Might as well try it |
|
Done. Build without cache running now. |
|
Well the |
|
Argh, still only restored the ones from before: |
|
I opened a forum topic: https://discuss.circleci.com/t/cache-not-restoring-properly/4896 |
I think how it works is that it uses the previous cache from the same branch. Only for new branches it uses the master's cache. I'll try with a new branch and see how it looks. |
|
And looks like for PRs it is using the master of the fork, so basically we'll see if clearing the cache worked when we merge something. |
|
So the problem with this might be that every PR is using the cache of the user instead of upstream master. In that case, I think this would be reasonable if we could somehow only download the data that is being used. |
|
Actually, in that case, we just won't auto-download anything. Each example when it runs will trigger a download if necessary. Let's try it. |
|
But the caching happens before the testing. So if we download at testing stage nothing gets cached. So at least for the master, the downloads should stay. |
Yeah -- check the commit I pushed, this is what happens. For |
Current coverage is 86.67%@@ master #3365 diff @@
==========================================
Files 335 335
Lines 57628 57628
Methods 0 0
Messages 0 0
Branches 8721 8781 +60
==========================================
+ Hits 30105 49948 +19843
+ Misses 25694 5004 -20690
- Partials 1829 2676 +847
|
|
Hooray, it worked! https://circleci.com/gh/mne-tools/mne-python/1446#artifacts @jaeilepp feel free to merge if you're happy. |
|
(I killed the Travis build manually since the last commit only changed Circle, look one commit back to see Travis passing) |
|
thx @Eric89GXL |
* 'master' of git://github.com/mne-tools/mne-python: (48 commits) FIX: Flake fix pep8 [MRG] FIX Topographic plotting for KIT-UMD data (mne-tools#3349) QUICKFIX: logging test Error message for simulate evoked. (mne-tools#3372) Warnings for big bems. (mne-tools#3374) Add test for baseline correction. Use apply_baseline in read_evokeds. Makes (None, 0) the default value baseline in apply_baseline. [FIX] Adds apply_baseline to Evoked fix component inds in ica tutorial (mne-tools#3379) FIX: Xdawn with shuffled epochs (mne-tools#3373) FIX consistency: Epochs.load_data() should always return self (mne-tools#3376) MRG: Build relevant examples (mne-tools#3365) [MRG] set vmin, vmax after smoothing in plot_epochs_image (mne-tools#3360) [FIX] Attempt to fix circle. fixing a bg plotting bug for topo plots Modified test and tutorial. [MRG] convert surface using python. (mne-tools#3273) [MRG] raw.plot_psd() with flat channel (mne-tools#3342) ...
4 participants
Closes #3357.
Today is Sunday right @agramfort? :)