[WIP] Add support for Studies

[janvanrijn]

Reference Issue

What does this PR implement/fix? Explain your changes.

How should this PR be tested?

Any other comments?

[codecov-io]

Codecov Report

Merging #620 into develop will increase coverage by 0.29%.
The diff coverage is 96.99%.

@@ Coverage Diff @@## develop #620 +/- ##
===========================================
+ Coverage 89.59% 89.88% +0.29% 
===========================================
Files 32 32 Lines 3123 3193 +70 ===========================================
+ Hits 2798 2870 +72 + Misses 325 323 -2

Impacted Files	Coverage Δ
openml/tasks/task.py	95.87% <100%> (ø)	⬆️
openml/study/study.py	100% <100%> (ø)	⬆️
openml/runs/functions.py	87.06% <100%> (+0.65%)	⬆️
openml/tasks/functions.py	88.38% <100%> (+0.07%)	⬆️
openml/flows/functions.py	93.25% <100%> (ø)	⬆️
openml/flows/flow.py	93.82% <100%> (ø)	⬆️
openml/datasets/dataset.py	77.53% <100%> (ø)	⬆️
openml/runs/run.py	89.01% <100%> (+0.04%)	⬆️
openml/evaluations/functions.py	95.55% <100%> (ø)	⬆️
openml/study/__init__.py	100% <100%> (ø)	⬆️
... and 6 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 96db525...bb0161a. Read the comment docs.

[mfeurer]

Looks good to me. Except that one line which basically says to not merge this PR.

[mfeurer]

And it seems there are some minor flake8 errors in the push test. I really don't know what to do with the flake8 errors in the PR test as they are so many and would only happen after the merge.

[janvanrijn]

Looks good to me. Except that one line which basically says to not merge this PR.

we need to wait for the PHP PR to be merged.

And it seems there are some minor flake8 errors in the push test. I really don't know what to do with the flake8 errors in the PR test as they are so many and would only happen after the merge.

Bigger issue: some lines are OK, other still contain mistakes (legacy mistakes, and we don't really want to solve them all at once)

Clearly it would be beneficial to have the whole project flake compatible. We can maybe ask the programmer to do so on one of the first days? (Once all big PR's are merged) From there on flake can run on all lines, not just the diff, and we don't have these problems anymore.

The problems introduced in this PR will be fixed by me.

[janvanrijn]

openml/study/__init__.py:2:1: F401 '.functions.get_study' imported but unused
from .functions import get_study, create_study, create_benchmark_suite, \
^
openml/study/__init__.py:2:1: F401 '.functions.create_study' imported but unused
from .functions import get_study, create_study, create_benchmark_suite, \
^
openml/study/__init__.py:2:1: F401 '.functions.create_benchmark_suite' imported but unused
from .functions import get_study, create_study, create_benchmark_suite, \
^
openml/study/__init__.py:2:1: F401 '.functions.delete_study' imported but unused
from .functions import get_study, create_study, create_benchmark_suite, \
^
openml/study/__init__.py:2:1: F401 '.functions.attach_to_study' imported but unused
from .functions import get_study, create_study, create_benchmark_suite, \
^
openml/study/__init__.py:2:1: F401 '.functions.detach_from_study' imported but unused
from .functions import get_study, create_study, create_benchmark_suite, \

flake8 residuals. this is fine right? Can we inform it that this warning in __init__ files is OK?

[mfeurer]

Could you please remove this call to the print function?

[mfeurer]

Looks good to me except for the failing unit tests.