Error in MFPBenchObjetiveFunction

[LukasFehring]

In the creation of MFPBenchObjetiveFunction, the value metrics is assigned a metric.

class MFPBenchObjectiveFunction(ObjectiveFunction):
    """MF-Prior-Bench ObjectiveFunction class."""

    def __init__(
        self,
        benchmark_name: str,
        metric: str | list[str],
        benchmark: str | None = None,
        budget_type: str | None = None,
        prior: str | Path | C | Mapping[str, Any] | None = None,
        perturb_prior: float | None = None,
        benchmark_kwargs: dict | None = None,
        loggers: list[AbstractLogger] | None = None,
    ) -> None:
        """Initialize a MF-Prior-Bench objective function."""
        super().__init__(loggers)

        self.benchmark_name = benchmark_name
        self.budget_type = budget_type
        self.benchmark = benchmark
        self.metrics = metric
        ...

If a string is passed as a metric, e.g. "valid_error_rate" it is utilized as a list of metrics ["v","a",...] causing errors in the evaluation.

---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
Cell In[7], [line 8](vscode-notebook-cell:?execution_count=7&line=8)
      5 baseline = obj.configspace.get_default_configuration()
      7 print("Create explanation task and instantiate HyperSHAP")
----> [8](vscode-notebook-cell:?execution_count=7&line=8) hypershap = HyperSHAP(explanation_task_from_carps_objective_function(obj))
     10 print("Tunability")
     11 hypershap.tunability(baseline)

Cell In[2], [line 6](vscode-notebook-cell:?execution_count=2&line=6)
      4 def explanation_task_from_carps_objective_function(objective_function: ObjectiveFunction) -> ExplanationTask:
      5     wrapper = ObjectiveFunctionWrapper(objective_function)
----> [6](vscode-notebook-cell:?execution_count=2&line=6)     return ExplanationTask.from_function(
      7         config_space=objective_function.configspace, function=wrapper.evaluate, n_samples=10_000
      8     )

File ~/Desktop/MergePriorLocations/.venv/lib/python3.13/site-packages/hypershap/task.py:149, in ExplanationTask.from_function(config_space, function, n_samples, base_model)
    136 """Create an ExplanationTask from a function that evaluates configurations.
    137 
    138 Args:
   (...)    146 
    147 """
    148 samples: list[Configuration] = config_space.sample_configuration(n_samples)
--> [149](https://file+.vscode-resource.vscode-cdn.net/Users/lukasfehring/Desktop/MergePriorLocations/~/Desktop/MergePriorLocations/.venv/lib/python3.13/site-packages/hypershap/task.py:149) values: list[float] = [function(config) for config in samples]
    150 data: list[tuple[Configuration, float]] = list(zip(samples, values, strict=False))
...
--> [139](https://file+.vscode-resource.vscode-cdn.net/Users/lukasfehring/Desktop/MergePriorLocations/~/Desktop/MergePriorLocations/.venv/lib/python3.13/site-packages/carps/objective_functions/mfpbench.py:139) ret = [result[metric] for metric in self.metrics]
    140 if len(ret) == 1:
    141     ret = ret[0]

KeyError: 'v'