FIX assign study's id to study_id instead of id

[PGijsbers]

A study's id would be assigned to id on initialization, and to study_id after publish. Now it is uniformly always study_id, following the other openml-python objects (e.g. run.run_id).

[PGijsbers]

Restarting the scikit-learn 0.21 test after getting an (what I think is) unrelated error:

=================================== FAILURES ===================================

___________________________ TestTask.test__get_task ____________________________

[gw2] linux -- Python 3.7.4 /home/travis/miniconda/envs/testenv/bin/python
self = <tests.test_tasks.test_task_functions.TestTask testMethod=test__get_task>
    def test__get_task(self):
        openml.config.cache_directory = self.static_cache_dir
>       openml.tasks.get_task(1882)

/home/travis/build/openml/openml-python/tests/test_tasks/test_task_functions.py:123: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

/home/travis/build/openml/openml-python/openml/utils.py:328: in safe_func
    return func(*args, **kwargs)
/home/travis/build/openml/openml-python/openml/tasks/functions.py:378: in get_task
    raise e
/home/travis/build/openml/openml-python/openml/tasks/functions.py:372: in get_task
    task.download_split()
/home/travis/build/openml/openml-python/openml/tasks/task.py:132: in download_split
    split = OpenMLSplit._from_arff_file(cached_split_file)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

cls = <class 'openml.tasks.split.OpenMLSplit'>
filename = '/home/travis/build/openml/openml-python/tests/files/org/openml/test/tasks1882/datasplits.arff'
    @classmethod

    def _from_arff_file(cls, filename: str) -> 'OpenMLSplit':  
        repetitions = None
        pkl_filename = filename.replace(".arff", ".pkl.py3")
        if os.path.exists(pkl_filename):
            with open(pkl_filename, "rb") as fh:
>               _ = pickle.load(fh)
E               EOFError: Ran out of input

/home/travis/build/openml/openml-python/openml/tasks/split.py:77: EOFError

___________________________ OpenMLSplitTest.test_eq ____________________________

[gw0] linux -- Python 3.7.4 /home/travis/miniconda/envs/testenv/bin/python
self = <tests.test_tasks.test_split.OpenMLSplitTest testMethod=test_eq>
    def test_eq(self):
        split = OpenMLSplit._from_arff_file(self.arff_filename)
        self.assertEqual(split, split) 
>       split2 = OpenMLSplit._from_arff_file(self.arff_filename)

/home/travis/build/openml/openml-python/tests/test_tasks/test_split.py:35: 

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

cls = <class 'openml.tasks.split.OpenMLSplit'>
filename = '/home/travis/build/openml/openml-python/tests/test_tasks/../files/org/openml/test/tasks/1882/datasplits.arff'
    @classmethod
    def _from_arff_file(cls, filename: str) -> 'OpenMLSplit':
        repetitions = None
        pkl_filename = filename.replace(".arff", ".pkl.py3")
        if os.path.exists(pkl_filename):
            with open(pkl_filename, "rb") as fh:
                _ = pickle.load(fh)
            repetitions = _["repetitions"]
            name = _["name"]    

        # Cache miss
        if repetitions is None:
            # Faster than liac-arff and sufficient in this situation!
            if not os.path.exists(filename):
                raise FileNotFoundError(
>                   'Split arff %s does not exist!' % filename
                )
E               FileNotFoundError: Split arff /home/travis/build/openml/openml-python/tests/test_tasks/../files/org/openml/test/tasks/1882/datasplits.arff does not exist!

/home/travis/build/openml/openml-python/openml/tasks/split.py:86: FileNotFoundError

_____________________ OpenMLSplitTest.test_from_arff_file ______________________

[gw0] linux -- Python 3.7.4 /home/travis/miniconda/envs/testenv/bin/python
self = <tests.test_tasks.test_split.OpenMLSplitTest testMethod=test_from_arff_file>
    def test_from_arff_file(self):
>       split = OpenMLSplit._from_arff_file(self.arff_filename)
/home/travis/build/openml/openml-python/tests/test_tasks/test_split.py:52: 

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

cls = <class 'openml.tasks.split.OpenMLSplit'>

filename = '/home/travis/build/openml/openml-python/tests/test_tasks/../files/org/openml/test/tasks/1882/datasplits.arff'

    @classmethod

    def _from_arff_file(cls, filename: str) -> 'OpenMLSplit':

    

        repetitions = None

    

        pkl_filename = filename.replace(".arff", ".pkl.py3")

    

        if os.path.exists(pkl_filename):

            with open(pkl_filename, "rb") as fh:

                _ = pickle.load(fh)

            repetitions = _["repetitions"]

            name = _["name"]

    

        # Cache miss

        if repetitions is None:

            # Faster than liac-arff and sufficient in this situation!

            if not os.path.exists(filename):

                raise FileNotFoundError(

>                   'Split arff %s does not exist!' % filename

                )

E               FileNotFoundError: Split arff /home/travis/build/openml/openml-python/tests/test_tasks/../files/org/openml/test/tasks/1882/datasplits.arff does not exist!

/home/travis/build/openml/openml-python/openml/tasks/split.py:86: FileNotFoundError

_________ OpenMLTaskMethodsTest.test_get_train_and_test_split_indices __________

[gw2] linux -- Python 3.7.4 /home/travis/miniconda/envs/testenv/bin/python

tid = 1882

    def _get_cached_task(tid: int) -> OpenMLTask:

        """Return a cached task based on the given id.

    

        Parameters

        ----------

        tid : int

            Id of the task.

    

        Returns

        -------

        OpenMLTask

        """

        tid_cache_dir = openml.utils._create_cache_directory_for_id(

            TASKS_CACHE_DIR_NAME,

            tid

        )

    

        try:

>           with io.open(os.path.join(tid_cache_dir, "task.xml"), encoding='utf8')\

                    as fh:

E                   FileNotFoundError: [Errno 2] No such file or directory: '/home/travis/build/openml/openml-python/tests/files/org/openml/test/tasks/1882/task.xml'

/home/travis/build/openml/openml-python/openml/tasks/functions.py:72: FileNotFoundError

During handling of the above exception, another exception occurred:

task_id = 1882

    def _get_task_description(task_id):

    

        try:

>           return _get_cached_task(task_id)

/home/travis/build/openml/openml-python/openml/tasks/functions.py:386: 

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

tid = 1882

    def _get_cached_task(tid: int) -> OpenMLTask:

        """Return a cached task based on the given id.

    

        Parameters

        ----------

        tid : int

            Id of the task.

    

        Returns

        -------

        OpenMLTask

        """

        tid_cache_dir = openml.utils._create_cache_directory_for_id(

            TASKS_CACHE_DIR_NAME,

            tid

        )

    

        try:

            with io.open(os.path.join(tid_cache_dir, "task.xml"), encoding='utf8')\

                    as fh:

                return _create_task_from_xml(fh.read())

        except (OSError, IOError):

            openml.utils._remove_cache_dir_for_id(TASKS_CACHE_DIR_NAME,

                                                  tid_cache_dir)

            raise OpenMLCacheException("Task file for tid %d not "

>                                      "cached" % tid)

E           openml.exceptions.OpenMLCacheException: Task file for tid 1882 not cached

/home/travis/build/openml/openml-python/openml/tasks/functions.py:79: OpenMLCacheException

During handling of the above exception, another exception occurred:

self = <tests.test_tasks.test_task_methods.OpenMLTaskMethodsTest testMethod=test_get_train_and_test_split_indices>

    def test_get_train_and_test_split_indices(self):

        openml.config.cache_directory = self.static_cache_dir

>       task = openml.tasks.get_task(1882)

/home/travis/build/openml/openml-python/tests/test_tasks/test_task_methods.py:31: 

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

/home/travis/build/openml/openml-python/openml/utils.py:328: in safe_func

    return func(*args, **kwargs)

/home/travis/build/openml/openml-python/openml/tasks/functions.py:378: in get_task

    raise e

/home/travis/build/openml/openml-python/openml/tasks/functions.py:360: in get_task

    task = _get_task_description(task_id)

/home/travis/build/openml/openml-python/openml/tasks/functions.py:396: in _get_task_description

    'get')

/home/travis/build/openml/openml-python/openml/_api_calls.py:51: in _perform_api_call

    return _read_url(url, request_method, data)

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

url = 'https://test.openml.org/api/v1/xml/task/1882', request_method = 'get'

data = {'api_key': '610344db6388d9ba34f6db45a3cf71de'}

    def _read_url(url, request_method, data=None):

        data = {} if data is None else data

        if config.apikey is not None:

            data['api_key'] = config.apikey

    

        response = send_request(request_method=request_method, url=url, data=data)

        if response.status_code != 200:

>           raise _parse_server_exception(response, url)

E           openml.exceptions.OpenMLServerException: Unknown task

/home/travis/build/openml/openml-python/openml/_api_calls.py:98: OpenMLServerException

[codecov-io]

Codecov Report

Merging #782 into develop will decrease coverage by 0.02%.
The diff coverage is 25%.

@@             Coverage Diff             @@
##           develop     #782      +/-   ##
===========================================
- Coverage    87.71%   87.69%   -0.03%     
===========================================
  Files           36       36              
  Lines         4177     4177              
===========================================
- Hits          3664     3663       -1     
- Misses         513      514       +1

Impacted Files	Coverage Δ
openml/study/study.py	64.36% <25%> (ø)	⬆️
openml/_api_calls.py	83.11% <0%> (-1.3%)	⬇️

---

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 afc7445...c046f59. Read the comment docs.