# Copyright 2020 The HuggingFace Datasets Authors and the current dataset script contributor.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""EER compute script"""

import evaluate
import datasets

import numpy as np 
import sklearn.metrics

# TODO: Add BibTeX citation
_CITATION = """\
@InProceedings{huggingface:module,
title = {A great new module},
authors={huggingface, Inc.},
year={2020}
}
"""

_DESCRIPTION = """\
This module is designed to compute Equal Error Rate metric, which is used a lot in the Automatic Speaker Verification task.
"""


_KWARGS_DESCRIPTION = """
The EER is the location on a ROC or DET curve where the false acceptance rate and false rejection rate are equal. In general, the lower the equal error rate value, the higher the accuracy of the biometric system.
"""


@evaluate.utils.file_utils.add_start_docstrings(_DESCRIPTION, _KWARGS_DESCRIPTION)
class EER(evaluate.Metric):
    """Compute Equal error rate metrics"""

    def _info(self):
        return evaluate.MetricInfo(
            module_type="metric",
            description=_DESCRIPTION,
            citation=_CITATION,
            inputs_description=_KWARGS_DESCRIPTION,
            features=datasets.Features({
                'predictions': datasets.Value('int64'),
                'references': datasets.Value('int64'),
                'pos_label': datasets.Value('int64')
            }),
            reference_urls=["https://github.com/YuanGongND/python-compute-eer"]
        )

    def _compute(self, predictions, references, pos_label=1):
        """Returns EER the scores"""
  
        fpr, tpr, threshold = sklearn.metrics.roc_curve(references, predictions, pos_label=pos_label)
        fnr = 1 - tpr

        # the threshold of fnr == fpr
        eer_threshold = threshold[np.nanargmin(np.absolute((fnr - fpr)))]

        # theoretically eer from fpr and eer from fnr should be identical but they can be slightly differ in reality
        eer_1 = fpr[np.nanargmin(np.absolute((fnr - fpr)))]
        eer_2 = fnr[np.nanargmin(np.absolute((fnr - fpr)))]

        # return the mean of eer from fpr and from fnr
        eer = (eer_1 + eer_2) / 2
  
        
        return {"eer": eer}