[BUG] create_task() hardcodes task_type="None" instead of inferring type string from TaskType #1766

return task_cls(
    task_id=None,
    task_type_id=task_type,
    task_type="None",  # TODO: refactor to get task type string from ID.
    data_set_id=dataset_id,
    target_name=target_name,
    estimation_procedure_id=estimation_procedure_id,
    evaluation_measure=evaluation_measure,
    **kwargs,
)
Because of this, task.task_type is literally the string "None" rather than the canonical task type name (e.g., "Supervised Classification", "Supervised Regression", "Clustering", "Learning Curve").

This causes:

Equality checks like if task.task_type == "Supervised Classification" (used in client code and in tests/test_runs/test_run.py:290) to evaluate to False.
Error messages and representations to display Task type 'None' (such as in openml/tasks/task.py:315).
Steps/Code to Reproduce
python
import openml
from openml.tasks import TaskType, create_task
task = create_task(
    TaskType.SUPERVISED_CLASSIFICATION,
    dataset_id=1,
    target_name="class",
    estimation_procedure_id=1,
)
print("task.task_type:", repr(task.task_type))
print("Is Supervised Classification?", task.task_type == "Supervised Classification")
Expected Results
task.task_type should be "Supervised Classification", and task.task_type == "Supervised Classification" should evaluate to True.

Actual Results
task.task_type: 'None'
Is Supervised Classification? False
Versions
OpenML: 0.16.0
Python: 3.10+
OS: Windows / Linux / macOS