-
Notifications
You must be signed in to change notification settings - Fork 388
/
f_beta.py
274 lines (221 loc) · 12.6 KB
/
f_beta.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
# Copyright The PyTorch 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 Any, Dict, Optional
from torch import Tensor
from torchmetrics.classification.stat_scores import StatScores
from torchmetrics.functional.classification.f_beta import _fbeta_compute
from torchmetrics.utilities.enums import AverageMethod
class FBetaScore(StatScores):
r"""Computes `F-score`_, specifically:
.. math::
F_\beta = (1 + \beta^2) * \frac{\text{precision} * \text{recall}}
{(\beta^2 * \text{precision}) + \text{recall}}
Where :math:`\beta` is some positive real factor. Works with binary, multiclass, and multilabel data.
Accepts logit scores or probabilities from a model output or integer class values in prediction.
Works with multi-dimensional preds and target.
Forward accepts
- ``preds`` (float or long tensor): ``(N, ...)`` or ``(N, C, ...)`` where C is the number of classes
- ``target`` (long tensor): ``(N, ...)``
If preds and target are the same shape and preds is a float tensor, we use the ``self.threshold`` argument
to convert into integer labels. This is the case for binary and multi-label logits and probabilities.
If preds has an extra dimension as in the case of multi-class scores we perform an argmax on ``dim=1``.
Args:
num_classes: Number of classes. Necessary for ``'macro'``, ``'weighted'`` and ``None`` average methods.
beta: Beta coefficient in the F measure.
threshold:
Threshold for transforming probability or logit predictions to binary (0,1) predictions, in the case
of binary or multi-label inputs. Default value of 0.5 corresponds to input being probabilities.
average:
Defines the reduction that is applied. Should be one of the following:
- ``'micro'`` [default]: Calculate the metric globally, across all samples and classes.
- ``'macro'``: Calculate the metric for each class separately, and average the
metrics across classes (with equal weights for each class).
- ``'weighted'``: Calculate the metric for each class separately, and average the
metrics across classes, weighting each class by its support (``tp + fn``).
- ``'none'`` or ``None``: Calculate the metric for each class separately, and return
the metric for every class.
- ``'samples'``: Calculate the metric for each sample, and average the metrics
across samples (with equal weights for each sample).
.. note:: What is considered a sample in the multi-dimensional multi-class case
depends on the value of ``mdmc_average``.
.. note:: If ``'none'`` and a given class doesn't occur in the ``preds`` or ``target``,
the value for the class will be ``nan``.
mdmc_average:
Defines how averaging is done for multi-dimensional multi-class inputs (on top of the
``average`` parameter). Should be one of the following:
- ``None`` [default]: Should be left unchanged if your data is not multi-dimensional
multi-class.
- ``'samplewise'``: In this case, the statistics are computed separately for each
sample on the ``N`` axis, and then averaged over samples.
The computation for each sample is done by treating the flattened extra axes ``...``
(see :ref:`pages/classification:input types`) as the ``N`` dimension within the sample,
and computing the metric for the sample based on that.
- ``'global'``: In this case the ``N`` and ``...`` dimensions of the inputs
(see :ref:`pages/classification:input types`)
are flattened into a new ``N_X`` sample axis, i.e. the inputs are treated as if they
were ``(N_X, C)``. From here on the ``average`` parameter applies as usual.
ignore_index:
Integer specifying a target class to ignore. If given, this class index does not contribute
to the returned score, regardless of reduction method. If an index is ignored, and ``average=None``
or ``'none'``, the score for the ignored class will be returned as ``nan``.
top_k:
Number of the highest probability or logit score predictions considered finding the correct label,
relevant only for (multi-dimensional) multi-class inputs. The default value (``None``) will be interpreted
as 1 for these inputs.
Should be left at default (``None``) for all other types of inputs.
multiclass:
Used only in certain special cases, where you want to treat inputs as a different type
than what they appear to be. See the parameter's
:ref:`documentation section <pages/classification:using the multiclass parameter>`
for a more detailed explanation and examples.
kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info.
Raises:
ValueError:
If ``average`` is none of ``"micro"``, ``"macro"``, ``"weighted"``, ``"none"``, ``None``.
Example:
>>> import torch
>>> from torchmetrics import FBetaScore
>>> target = torch.tensor([0, 1, 2, 0, 1, 2])
>>> preds = torch.tensor([0, 2, 1, 0, 0, 1])
>>> f_beta = FBetaScore(num_classes=3, beta=0.5)
>>> f_beta(preds, target)
tensor(0.3333)
"""
full_state_update: bool = False
def __init__(
self,
num_classes: Optional[int] = None,
beta: float = 1.0,
threshold: float = 0.5,
average: str = "micro",
mdmc_average: Optional[str] = None,
ignore_index: Optional[int] = None,
top_k: Optional[int] = None,
multiclass: Optional[bool] = None,
**kwargs: Dict[str, Any],
) -> None:
self.beta = beta
allowed_average = list(AverageMethod)
if average not in allowed_average:
raise ValueError(f"The `average` has to be one of {allowed_average}, got {average}.")
if "reduce" not in kwargs or kwargs["reduce"] is None:
kwargs["reduce"] = "macro" if average in [AverageMethod.WEIGHTED, AverageMethod.NONE, "none"] else average
if "mdmc_reduce" not in kwargs or kwargs["mdmc_reduce"] is None:
kwargs["mdmc_reduce"] = mdmc_average
super().__init__(
threshold=threshold,
top_k=top_k,
num_classes=num_classes,
multiclass=multiclass,
ignore_index=ignore_index,
**kwargs,
)
self.average = average
def compute(self) -> Tensor:
"""Computes f-beta over state."""
tp, fp, tn, fn = self._get_final_stats()
return _fbeta_compute(tp, fp, tn, fn, self.beta, self.ignore_index, self.average, self.mdmc_reduce)
class F1Score(FBetaScore):
"""Computes F1 metric.
F1 metrics correspond to a harmonic mean of the precision and recall scores.
Works with binary, multiclass, and multilabel data. Accepts logits or probabilities from a model
output or integer class values in prediction. Works with multi-dimensional preds and target.
Forward accepts
- ``preds`` (float or long tensor): ``(N, ...)`` or ``(N, C, ...)`` where C is the number of classes
- ``target`` (long tensor): ``(N, ...)``
If preds and target are the same shape and preds is a float tensor, we use the ``self.threshold`` argument.
This is the case for binary and multi-label logits.
If preds has an extra dimension as in the case of multi-class scores we perform an argmax on ``dim=1``.
Args:
num_classes:
Number of classes. Necessary for ``'macro'``, ``'weighted'`` and ``None`` average methods.
threshold:
Threshold for transforming probability or logit predictions to binary ``(0,1)`` predictions, in the case
of binary or multi-label inputs. Default value of ``0.5`` corresponds to input being probabilities.
average:
Defines the reduction that is applied. Should be one of the following:
- ``'micro'`` [default]: Calculate the metric globally, across all samples and classes.
- ``'macro'``: Calculate the metric for each class separately, and average the
metrics across classes (with equal weights for each class).
- ``'weighted'``: Calculate the metric for each class separately, and average the
metrics across classes, weighting each class by its support (``tp + fn``).
- ``'none'`` or ``None``: Calculate the metric for each class separately, and return
the metric for every class.
- ``'samples'``: Calculate the metric for each sample, and average the metrics
across samples (with equal weights for each sample).
.. note:: What is considered a sample in the multi-dimensional multi-class case
depends on the value of ``mdmc_average``.
mdmc_average:
Defines how averaging is done for multi-dimensional multi-class inputs (on top of the
``average`` parameter). Should be one of the following:
- ``None`` [default]: Should be left unchanged if your data is not multi-dimensional
multi-class.
- ``'samplewise'``: In this case, the statistics are computed separately for each
sample on the ``N`` axis, and then averaged over samples.
The computation for each sample is done by treating the flattened extra axes ``...``
(see :ref:`pages/classification:input types`) as the ``N`` dimension within the sample,
and computing the metric for the sample based on that.
- ``'global'``: In this case the ``N`` and ``...`` dimensions of the inputs
(see :ref:`pages/classification:input types`)
are flattened into a new ``N_X`` sample axis, i.e. the inputs are treated as if they
were ``(N_X, C)``. From here on the ``average`` parameter applies as usual.
ignore_index:
Integer specifying a target class to ignore. If given, this class index does not contribute
to the returned score, regardless of reduction method. If an index is ignored, and ``average=None``
or ``'none'``, the score for the ignored class will be returned as ``nan``.
top_k:
Number of the highest probability or logit score predictions considered finding the correct label,
relevant only for (multi-dimensional) multi-class inputs. The
default value (``None``) will be interpreted as 1 for these inputs.
Should be left at default (``None``) for all other types of inputs.
multiclass:
Used only in certain special cases, where you want to treat inputs as a different type
than what they appear to be. See the parameter's
:ref:`documentation section <pages/classification:using the multiclass parameter>`
for a more detailed explanation and examples.
kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info.
Example:
>>> import torch
>>> from torchmetrics import F1Score
>>> target = torch.tensor([0, 1, 2, 0, 1, 2])
>>> preds = torch.tensor([0, 2, 1, 0, 0, 1])
>>> f1 = F1Score(num_classes=3)
>>> f1(preds, target)
tensor(0.3333)
"""
is_differentiable: bool = False
higher_is_better: bool = True
full_state_update: bool = False
def __init__(
self,
num_classes: Optional[int] = None,
threshold: float = 0.5,
average: str = "micro",
mdmc_average: Optional[str] = None,
ignore_index: Optional[int] = None,
top_k: Optional[int] = None,
multiclass: Optional[bool] = None,
**kwargs: Dict[str, Any],
) -> None:
super().__init__(
num_classes=num_classes,
beta=1.0,
threshold=threshold,
average=average,
mdmc_average=mdmc_average,
ignore_index=ignore_index,
top_k=top_k,
multiclass=multiclass,
**kwargs,
)