|
1 | | -from typing import Callable, Union |
2 | | - |
3 | | -import torch |
4 | | - |
5 | | -from ignite.metrics import EpochMetric |
6 | | - |
7 | | - |
8 | | -def average_precision_compute_fn(y_preds: torch.Tensor, y_targets: torch.Tensor) -> float: |
9 | | - from sklearn.metrics import average_precision_score |
10 | | - |
11 | | - y_true = y_targets.cpu().numpy() |
12 | | - y_pred = y_preds.cpu().numpy() |
13 | | - return average_precision_score(y_true, y_pred) |
14 | | - |
15 | | - |
16 | | -class AveragePrecision(EpochMetric): |
17 | | - """Computes Average Precision accumulating predictions and the ground-truth during an epoch |
18 | | - and applying `sklearn.metrics.average_precision_score <https://scikit-learn.org/stable/modules/generated/ |
19 | | - sklearn.metrics.average_precision_score.html#sklearn.metrics.average_precision_score>`_ . |
20 | | -
|
21 | | - Args: |
22 | | - output_transform: a callable that is used to transform the |
23 | | - :class:`~ignite.engine.engine.Engine`'s ``process_function``'s output into the |
24 | | - form expected by the metric. This can be useful if, for example, you have a multi-output model and |
25 | | - you want to compute the metric with respect to one of the outputs. |
26 | | - check_compute_fn: Default False. If True, `average_precision_score |
27 | | - <https://scikit-learn.org/stable/modules/generated/sklearn.metrics.average_precision_score.html |
28 | | - #sklearn.metrics.average_precision_score>`_ is run on the first batch of data to ensure there are |
29 | | - no issues. User will be warned in case there are any issues computing the function. |
30 | | - device: optional device specification for internal storage. |
31 | | -
|
32 | | - Note: |
33 | | - AveragePrecision expects y to be comprised of 0's and 1's. y_pred must either be probability estimates or |
34 | | - confidence values. To apply an activation to y_pred, use output_transform as shown below: |
35 | | -
|
36 | | - .. code-block:: python |
37 | | -
|
38 | | - def activated_output_transform(output): |
39 | | - y_pred, y = output |
40 | | - y_pred = torch.softmax(y_pred, dim=1) |
41 | | - return y_pred, y |
42 | | - avg_precision = AveragePrecision(activated_output_transform) |
43 | | -
|
44 | | - Examples: |
45 | | -
|
46 | | - .. include:: defaults.rst |
47 | | - :start-after: :orphan: |
48 | | -
|
49 | | - .. testcode:: |
50 | | -
|
51 | | - y_pred = torch.tensor([[0.79, 0.21], [0.30, 0.70], [0.46, 0.54], [0.16, 0.84]]) |
52 | | - y_true = torch.tensor([[1, 1], [1, 1], [0, 1], [0, 1]]) |
53 | | -
|
54 | | - avg_precision = AveragePrecision() |
55 | | - avg_precision.attach(default_evaluator, 'average_precision') |
56 | | - state = default_evaluator.run([[y_pred, y_true]]) |
57 | | - print(state.metrics['average_precision']) |
58 | | -
|
59 | | - .. testoutput:: |
60 | | -
|
61 | | - 0.9166... |
62 | | -
|
63 | | - """ |
64 | | - |
65 | | - def __init__( |
66 | | - self, |
67 | | - output_transform: Callable = lambda x: x, |
68 | | - check_compute_fn: bool = False, |
69 | | - device: Union[str, torch.device] = torch.device("cpu"), |
70 | | - ): |
71 | | - try: |
72 | | - from sklearn.metrics import average_precision_score # noqa: F401 |
73 | | - except ImportError: |
74 | | - raise ModuleNotFoundError("This contrib module requires scikit-learn to be installed.") |
75 | | - |
76 | | - super(AveragePrecision, self).__init__( |
77 | | - average_precision_compute_fn, |
78 | | - output_transform=output_transform, |
79 | | - check_compute_fn=check_compute_fn, |
80 | | - device=device, |
81 | | - ) |
| 1 | +""" ``ignite.contrib.metrics.average_precision`` was moved to ``ignite.metrics.average_precision``. |
| 2 | +Note: |
| 3 | + ``ignite.contrib.metrics.average_precision`` was moved to ``ignite.metrics.average_precision``. |
| 4 | + Please refer to :mod:`~ignite.metrics.average_precision`. |
| 5 | +""" |
| 6 | + |
| 7 | +import warnings |
| 8 | + |
| 9 | +removed_in = "0.6.0" |
| 10 | +deprecation_warning = ( |
| 11 | + f"{__file__} has been moved to /ignite/metrics/average_precision.py" |
| 12 | + + (f" and will be removed in version {removed_in}" if removed_in else "") |
| 13 | + + ".\n Please refer to the documentation for more details." |
| 14 | +) |
| 15 | +warnings.warn(deprecation_warning, DeprecationWarning, stacklevel=2) |
| 16 | +from ignite.metrics.average_precision import AveragePrecision |
| 17 | + |
| 18 | +__all__ = [ |
| 19 | + "AveragePrecision", |
| 20 | +] |
| 21 | + |
| 22 | +AveragePrecision = AveragePrecision |
0 commit comments