precision_fixed_recall.py

# Copyright The Lightning team.
#
# 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.
from typing import List, Optional, Tuple, Union

import torch
from torch import Tensor
from typing_extensions import Literal

from torchmetrics.functional.classification.precision_recall_curve import (
    _binary_precision_recall_curve_format,
    _binary_precision_recall_curve_tensor_validation,
    _binary_precision_recall_curve_update,
    _multiclass_precision_recall_curve_format,
    _multiclass_precision_recall_curve_tensor_validation,
    _multiclass_precision_recall_curve_update,
    _multilabel_precision_recall_curve_format,
    _multilabel_precision_recall_curve_tensor_validation,
    _multilabel_precision_recall_curve_update,
)
from torchmetrics.functional.classification.recall_fixed_precision import (
    _binary_recall_at_fixed_precision_arg_validation,
    _binary_recall_at_fixed_precision_compute,
    _multiclass_recall_at_fixed_precision_arg_compute,
    _multiclass_recall_at_fixed_precision_arg_validation,
    _multilabel_recall_at_fixed_precision_arg_compute,
    _multilabel_recall_at_fixed_precision_arg_validation,
)
from torchmetrics.utilities.enums import ClassificationTask


def _precision_at_recall(
    precision: Tensor,
    recall: Tensor,
    thresholds: Tensor,
    min_recall: float,
) -> Tuple[Tensor, Tensor]:
    try:
        max_precision, _, best_threshold = max(
            (p, r, t) for p, r, t in zip(precision, recall, thresholds) if r >= min_recall
        )

    except ValueError:
        max_precision = torch.tensor(0.0, device=precision.device, dtype=precision.dtype)
        best_threshold = torch.tensor(0)

    if max_precision == 0.0:
        best_threshold = torch.tensor(1e6, device=thresholds.device, dtype=thresholds.dtype)

    return max_precision, best_threshold


def binary_precision_at_fixed_recall(
    preds: Tensor,
    target: Tensor,
    min_recall: float,
    thresholds: Optional[Union[int, List[float], Tensor]] = None,
    ignore_index: Optional[int] = None,
    validate_args: bool = True,
) -> Tuple[Tensor, Tensor]:
    r"""Compute the highest possible precision value given the minimum recall thresholds provided for binary tasks.

    This is done by first calculating the precision-recall curve for different thresholds and the find the precision
    for a given recall level.

    Accepts the following input tensors:

    - ``preds`` (float tensor): ``(N, ...)``. Preds should be a tensor containing probabilities or logits for each
      observation. If preds has values outside [0,1] range we consider the input to be logits and will auto apply
      sigmoid per element.
    - ``target`` (int tensor): ``(N, ...)``. Target should be a tensor containing ground truth labels, and therefore
      only contain {0,1} values (except if `ignore_index` is specified). The value 1 always encodes the positive class.

    Additional dimension ``...`` will be flattened into the batch dimension.

    The implementation both supports calculating the metric in a non-binned but accurate version and a binned version
    that is less accurate but more memory efficient. Setting the `thresholds` argument to ``None`` will activate the
    non-binned  version that uses memory of size :math:`\mathcal{O}(n_{samples})` whereas setting the `thresholds`
    argument to either an integer, list or a 1d tensor will use a binned version that uses memory of
    size :math:`\mathcal{O}(n_{thresholds})` (constant memory).

    Args:
        preds: Tensor with predictions
        target: Tensor with true labels
        min_recall: float value specifying minimum recall threshold.
        thresholds:
            Can be one of:

            - If set to ``None``, will use a non-binned approach where thresholds are dynamically calculated from
              all the data. Most accurate but also most memory consuming approach.
            - If set to an ``int`` (larger than 1), will use that number of thresholds linearly spaced from
              0 to 1 as bins for the calculation.
            - If set to an ``list`` of floats, will use the indicated thresholds in the list as bins for the calculation
            - If set to an 1d :class:`~torch.Tensor` of floats, will use the indicated thresholds in the tensor as
              bins for the calculation.

        ignore_index:
            Specifies a target value that is ignored and does not contribute to the metric calculation
        validate_args: bool indicating if input arguments and tensors should be validated for correctness.
            Set to ``False`` for faster computations.

    Returns:
        (tuple): a tuple of 2 tensors containing:

        - precision: an scalar tensor with the maximum precision for the given precision level
        - threshold: an scalar tensor with the corresponding threshold level

    Example:
        >>> from torchmetrics.functional.classification import binary_precision_at_fixed_recall
        >>> preds = torch.tensor([0, 0.5, 0.7, 0.8])
        >>> target = torch.tensor([0, 1, 1, 0])
        >>> binary_precision_at_fixed_recall(preds, target, min_recall=0.5, thresholds=None)
        (tensor(0.6667), tensor(0.5000))
        >>> binary_precision_at_fixed_recall(preds, target, min_recall=0.5, thresholds=5)
        (tensor(0.6667), tensor(0.5000))
    """
    if validate_args:
        _binary_recall_at_fixed_precision_arg_validation(min_recall, thresholds, ignore_index)
        _binary_precision_recall_curve_tensor_validation(preds, target, ignore_index)
    preds, target, thresholds = _binary_precision_recall_curve_format(preds, target, thresholds, ignore_index)
    state = _binary_precision_recall_curve_update(preds, target, thresholds)
    return _binary_recall_at_fixed_precision_compute(
        state, thresholds, min_precision=min_recall, reduce_fn=_precision_at_recall
    )


def multiclass_precision_at_fixed_recall(
    preds: Tensor,
    target: Tensor,
    num_classes: int,
    min_recall: float,
    thresholds: Optional[Union[int, List[float], Tensor]] = None,
    ignore_index: Optional[int] = None,
    validate_args: bool = True,
) -> Tuple[Tensor, Tensor]:
    r"""Compute the highest possible precision value given the minimum recall thresholds provided for multiclass tasks.

    This is done by first calculating the precision-recall curve for different thresholds and the find the precision
    for a given recall level.

    Accepts the following input tensors:

    - ``preds`` (float tensor): ``(N, C, ...)``. Preds should be a tensor containing probabilities or logits for each
      observation. If preds has values outside [0,1] range we consider the input to be logits and will auto apply
      softmax per sample.
    - ``target`` (int tensor): ``(N, ...)``. Target should be a tensor containing ground truth labels, and therefore
      only contain values in the [0, n_classes-1] range (except if `ignore_index` is specified).

    Additional dimension ``...`` will be flattened into the batch dimension.

    The implementation both supports calculating the metric in a non-binned but accurate version and a binned version
    that is less accurate but more memory efficient. Setting the `thresholds` argument to ``None`` will activate the
    non-binned  version that uses memory of size :math:`\mathcal{O}(n_{samples})` whereas setting the `thresholds`
    argument to either an integer, list or a 1d tensor will use a binned version that uses memory of
    size :math:`\mathcal{O}(n_{thresholds} \times n_{classes})` (constant memory).

    Args:
        preds: Tensor with predictions
        target: Tensor with true labels
        num_classes: Integer specifing the number of classes
        min_recall: float value specifying minimum recall threshold.
        thresholds:
            Can be one of:

            - If set to ``None``, will use a non-binned approach where thresholds are dynamically calculated from
              all the data. Most accurate but also most memory consuming approach.
            - If set to an ``int`` (larger than 1), will use that number of thresholds linearly spaced from
              0 to 1 as bins for the calculation.
            - If set to an ``list`` of floats, will use the indicated thresholds in the list as bins for the calculation
            - If set to an 1d :class:`~torch.Tensor` of floats, will use the indicated thresholds in the tensor as
              bins for the calculation.

        ignore_index:
            Specifies a target value that is ignored and does not contribute to the metric calculation
        validate_args: bool indicating if input arguments and tensors should be validated for correctness.
            Set to ``False`` for faster computations.

    Returns:
        (tuple): a tuple of either 2 tensors or 2 lists containing

        - precision: an 1d tensor of size (n_classes, ) with the maximum precision for the given recall level per class
        - thresholds: an 1d tensor of size (n_classes, ) with the corresponding threshold level per class

    Example:
        >>> from torchmetrics.functional.classification import multiclass_precision_at_fixed_recall
        >>> preds = torch.tensor([[0.75, 0.05, 0.05, 0.05, 0.05],
        ...                       [0.05, 0.75, 0.05, 0.05, 0.05],
        ...                       [0.05, 0.05, 0.75, 0.05, 0.05],
        ...                       [0.05, 0.05, 0.05, 0.75, 0.05]])
        >>> target = torch.tensor([0, 1, 3, 2])
        >>> multiclass_precision_at_fixed_recall(  # doctest: +NORMALIZE_WHITESPACE
        ...     preds, target, num_classes=5, min_recall=0.5, thresholds=None)
        (tensor([1.0000, 1.0000, 0.2500, 0.2500, 0.0000]),
         tensor([7.5000e-01, 7.5000e-01, 5.0000e-02, 5.0000e-02, 1.0000e+06]))
        >>> multiclass_precision_at_fixed_recall(  # doctest: +NORMALIZE_WHITESPACE
        ...     preds, target, num_classes=5, min_recall=0.5, thresholds=5)
        (tensor([1.0000, 1.0000, 0.2500, 0.2500, 0.0000]),
         tensor([7.5000e-01, 7.5000e-01, 0.0000e+00, 0.0000e+00, 1.0000e+06]))
    """
    if validate_args:
        _multiclass_recall_at_fixed_precision_arg_validation(num_classes, min_recall, thresholds, ignore_index)
        _multiclass_precision_recall_curve_tensor_validation(preds, target, num_classes, ignore_index)
    preds, target, thresholds = _multiclass_precision_recall_curve_format(
        preds, target, num_classes, thresholds, ignore_index
    )
    state = _multiclass_precision_recall_curve_update(preds, target, num_classes, thresholds)
    return _multiclass_recall_at_fixed_precision_arg_compute(
        state, num_classes, thresholds, min_precision=min_recall, reduce_fn=_precision_at_recall
    )


def multilabel_precision_at_fixed_recall(
    preds: Tensor,
    target: Tensor,
    num_labels: int,
    min_recall: float,
    thresholds: Optional[Union[int, List[float], Tensor]] = None,
    ignore_index: Optional[int] = None,
    validate_args: bool = True,
) -> Tuple[Tensor, Tensor]:
    r"""Compute the highest possible precision value given the minimum recall thresholds provided for multilabel tasks.

    This is done by first calculating the precision-recall curve for different thresholds and the find the precision
    for a given recall level.

    Accepts the following input tensors:

    - ``preds`` (float tensor): ``(N, C, ...)``. Preds should be a tensor containing probabilities or logits for each
      observation. If preds has values outside [0,1] range we consider the input to be logits and will auto apply
      sigmoid per element.
    - ``target`` (int tensor): ``(N, C, ...)``. Target should be a tensor containing ground truth labels, and therefore
      only contain {0,1} values (except if `ignore_index` is specified).

    Additional dimension ``...`` will be flattened into the batch dimension.

    The implementation both supports calculating the metric in a non-binned but accurate version and a binned version
    that is less accurate but more memory efficient. Setting the `thresholds` argument to ``None`` will activate the
    non-binned  version that uses memory of size :math:`\mathcal{O}(n_{samples})` whereas setting the `thresholds`
    argument to either an integer, list or a 1d tensor will use a binned version that uses memory of
    size :math:`\mathcal{O}(n_{thresholds} \times n_{labels})` (constant memory).

    Args:
        preds: Tensor with predictions
        target: Tensor with true labels
        num_labels: Integer specifing the number of labels
        min_recall: float value specifying minimum recall threshold.
        thresholds:
            Can be one of:

            - If set to ``None``, will use a non-binned approach where thresholds are dynamically calculated from
              all the data. Most accurate but also most memory consuming approach.
            - If set to an ``int`` (larger than 1), will use that number of thresholds linearly spaced from
              0 to 1 as bins for the calculation.
            - If set to an ``list`` of floats, will use the indicated thresholds in the list as bins for the calculation
            - If set to an 1d :class:`~torch.Tensor` of floats, will use the indicated thresholds in the tensor as
              bins for the calculation.

        ignore_index:
            Specifies a target value that is ignored and does not contribute to the metric calculation
        validate_args: bool indicating if input arguments and tensors should be validated for correctness.
            Set to ``False`` for faster computations.

    Returns:
        (tuple): a tuple of either 2 tensors or 2 lists containing

        - precision: an 1d tensor of size (n_classes, ) with the maximum precision for the given recall level per class
        - thresholds: an 1d tensor of size (n_classes, ) with the corresponding threshold level per class

    Example:
        >>> from torchmetrics.functional.classification import multilabel_precision_at_fixed_recall
        >>> preds = torch.tensor([[0.75, 0.05, 0.35],
        ...                       [0.45, 0.75, 0.05],
        ...                       [0.05, 0.55, 0.75],
        ...                       [0.05, 0.65, 0.05]])
        >>> target = torch.tensor([[1, 0, 1],
        ...                        [0, 0, 0],
        ...                        [0, 1, 1],
        ...                        [1, 1, 1]])
        >>> multilabel_precision_at_fixed_recall(preds, target, num_labels=3, min_recall=0.5, thresholds=None)
        (tensor([1.0000, 0.6667, 1.0000]), tensor([0.7500, 0.5500, 0.3500]))
        >>> multilabel_precision_at_fixed_recall(preds, target, num_labels=3, min_recall=0.5, thresholds=5)
        (tensor([1.0000, 0.6667, 1.0000]), tensor([0.7500, 0.5000, 0.2500]))
    """
    if validate_args:
        _multilabel_recall_at_fixed_precision_arg_validation(num_labels, min_recall, thresholds, ignore_index)
        _multilabel_precision_recall_curve_tensor_validation(preds, target, num_labels, ignore_index)
    preds, target, thresholds = _multilabel_precision_recall_curve_format(
        preds, target, num_labels, thresholds, ignore_index
    )
    state = _multilabel_precision_recall_curve_update(preds, target, num_labels, thresholds)
    return _multilabel_recall_at_fixed_precision_arg_compute(
        state, num_labels, thresholds, ignore_index, min_precision=min_recall, reduce_fn=_precision_at_recall
    )


def precision_at_fixed_recall(
    preds: Tensor,
    target: Tensor,
    task: Literal["binary", "multiclass", "multilabel"],
    min_recall: float,
    thresholds: Optional[Union[int, List[float], Tensor]] = None,
    num_classes: Optional[int] = None,
    num_labels: Optional[int] = None,
    ignore_index: Optional[int] = None,
    validate_args: bool = True,
) -> Optional[Tuple[Tensor, Tensor]]:
    r"""Compute the highest possible recall value given the minimum precision thresholds provided.

    This is done by first calculating the precision-recall curve for different thresholds and the find the recall for a
    given precision level.

    This function is a simple wrapper to get the task specific versions of this metric, which is done by setting the
    ``task`` argument to either ``'binary'``, ``'multiclass'`` or ``multilabel``. See the documentation of
    :func:`binary_precision_at_fixed_recall`, :func:`multiclass_precision_at_fixed_recall` and
    :func:`multilabel_precision_at_fixed_recall` for the specific details of each argument influence and examples.
    """
    task = ClassificationTask.from_str(task)
    if task == ClassificationTask.BINARY:
        return binary_precision_at_fixed_recall(preds, target, min_recall, thresholds, ignore_index, validate_args)
    if task == ClassificationTask.MULTICLASS:
        if not isinstance(num_classes, int):
            raise ValueError(f"`num_classes` is expected to be `int` but `{type(num_classes)} was passed.`")
        return multiclass_precision_at_fixed_recall(
            preds, target, num_classes, min_recall, thresholds, ignore_index, validate_args
        )
    if task == ClassificationTask.MULTILABEL:
        if not isinstance(num_labels, int):
            raise ValueError(f"`num_labels` is expected to be `int` but `{type(num_labels)} was passed.`")
        return multilabel_precision_at_fixed_recall(
            preds, target, num_labels, min_recall, thresholds, ignore_index, validate_args
        )
    return None