[Bug]: sporadic multiobj optimization failure on boosting pipelines

[Lopa10ko]

Current Behavior

Getting ValueError: Length of values (0) does not match length of index (606) on boosting pipelines predict stage in run_classification_multiobj_example

Steps to Reproduce

2024-09-24_21-25-09_pipeline_saved.zip

import pandas as pd

from fedot import Fedot
from fedot.core.pipelines.pipeline import Pipeline
from fedot.core.utils import fedot_project_root


def run_classification_multiobj_example(visualization=False, timeout=1, with_tuning=True):
    train_data = pd.read_csv(f'{fedot_project_root()}/examples/data/Hill_Valley_with_noise_Training.data')
    test_data = pd.read_csv(f'{fedot_project_root()}/examples/data/Hill_Valley_with_noise_Testing.data')
    target = test_data['class']
    del test_data['class']
    problem = 'classification'

    metric_names = ['f1', 'node_number']
    initial_assumption = Pipeline().load('./2024-09-24_21-25-09_pipeline_saved')
    auto_model = Fedot(problem=problem, timeout=timeout, preset='best_quality',
                       metric=metric_names,
                       with_tuning=with_tuning,
                       initial_assumption=initial_assumption)
    auto_model.fit(features=train_data, target='class')
    prediction = auto_model.predict_proba(features=test_data)
    print(auto_model.get_metrics(target))

    if visualization:
        auto_model.plot_prediction()
        auto_model.plot_pareto()

    return prediction


if __name__ == '__main__':
    run_classification_multiobj_example(timeout=0.1, with_tuning=False)

Possible solution

Check whether data.target is set to an empty sequence:

FEDOT/fedot/core/operations/evaluation/operation_implementations/models/boostings_implementations.py

Lines 236 to 254 in 5be9119

	@staticmethod
	def convert_to_dataframe(data: Optional[InputData], identify_cats: bool):
	dataframe = pd.DataFrame(data=data.features, columns=data.features_names)
	if data.target is not None:
	dataframe['target'] = np.ravel(data.target)
	else:
	# TODO: temp workaround in case data.target is set to None intentionally
	# for test.integration.models.test_model.check_predict_correct
	dataframe['target'] = np.zeros(len(data.features))
	if identify_cats and data.categorical_idx is not None:
	for col in dataframe.columns[data.categorical_idx]:
	dataframe[col] = dataframe[col].astype('category')
	if data.numerical_idx is not None:
	for col in dataframe.columns[data.numerical_idx]:
	dataframe[col] = dataframe[col].astype('float')
	return dataframe.drop(columns=['target']), dataframe['target']

Context

failed integration tests run https://github.com/aimclub/FEDOT/actions/runs/11013973383
related to #1316