download errors, no kernel

[kegl]

There are weird errors in the full test: https://travis-ci.org/paris-saclay-cds/rampwf-kits-test-master

[kegl]

The storage is up, so I'm not sure what's happening. The titanic error is even stranger.

[jorisvandenbossche]

some first observations:

• The titanic one is an error with testing the notebook: it's again a problem with the notebook's kernel being python2 when ran from python 3. I think we have handled it before, but there might have been a change on the nbconvert side.
  This actually seems to have been failing for several months

• The mars craters one fails complaining about the data_train.npy and labels_train.csv not being found. @aboucaud any idea what is happening here? I don't think we changed the naming of the files in the summer ?

• The california one: FileNotFoundError: [Errno 2] No such file or directory: '/home/travis/build/paris-saclay-cds/rampwf-kits-test-master/california_rainfall/data/train_TS.nc'.
  This one also has been failing since the beginning.

[aboucaud]

The mars craters one fails complaining about the data_train.npy and labels_train.csv not being found. @aboucaud <https://github.com/aboucaud> any idea what is happening here? I don't think we changed the naming of the files in the summer ?

There is no sign of the `mars_craters` data directory on the ramp server, but there is `mars_craters_2`. I don't remember suppressing the first edition though. What shall we do then ?

[jorisvandenbossche]

But that is updated in https://github.com/ramp-kits/mars_craters/blob/master/download_data.py ?

[aboucaud]

then I have no clue..

…

Le 1 oct. 2018 à 15:17, Joris Van den Bossche ***@***.***> a écrit : But that is updated in https://github.com/ramp-kits/mars_craters/blob/master/download_data.py <https://github.com/ramp-kits/mars_craters/blob/master/download_data.py> ? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#151 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AC7F3qy54p7TcgsNy123jF2ReV5xoz_dks5ughX0gaJpZM4XB8nX>.

[kegl]

I think I deleted mars_craters for space. Then I discovered it was the DB :).

But this failure is not in the kit, rather the full test of ramp-workflow. Is that calling the kit's test? I don't remember.

[jorisvandenbossche]

But this failure is not in the kit, rather the full test of ramp-workflow. Is that calling the kit's test? I don't remember.

What do you mean with "the full test of ramp-workfow"?
What you linked above was https://travis-ci.org/paris-saclay-cds/rampwf-kits-test-master, which are the individual kit tests.

[kegl]

so then it's weird that the kit tests work but not https://travis-ci.org/paris-saclay-cds/rampwf-kits-test-master

[jorisvandenbossche]

OK, found the problem: with the changes in the summer, we didn't update the "testing" mode to keep it working.

On the kit's tests (https://travis-ci.org/ramp-kits/mars_craters), we actually download the full data, but on the cron job (https://travis-ci.org/paris-saclay-cds/rampwf-kits-test-master) we use the testing mode (os.environ['RAMP_TEST_MODE'] = '1')

[kegl]

It's still failing