-
Notifications
You must be signed in to change notification settings - Fork 387
/
homogeneity_completeness_v_measure.py
328 lines (248 loc) · 12.6 KB
/
homogeneity_completeness_v_measure.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
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
# 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 Any, List, Optional, Sequence, Union
from torch import Tensor
from torchmetrics.functional.clustering.homogeneity_completeness_v_measure import (
completeness_score,
homogeneity_score,
v_measure_score,
)
from torchmetrics.metric import Metric
from torchmetrics.utilities.data import dim_zero_cat
from torchmetrics.utilities.imports import _MATPLOTLIB_AVAILABLE
from torchmetrics.utilities.plot import _AX_TYPE, _PLOT_OUT_TYPE
if not _MATPLOTLIB_AVAILABLE:
__doctest_skip__ = ["HomogeneityScore.plot", "CompletenessScore.plot", "VMeasureScore.plot"]
class HomogeneityScore(Metric):
r"""Compute `Homogeneity Score`_.
The homogeneity score is a metric to measure the homogeneity of a clustering. A clustering result satisfies
homogeneity if all of its clusters contain only data points which are members of a single class. The metric is not
symmetric, therefore swapping ``preds`` and ``target`` yields a different score.
This clustering metric is an extrinsic measure, because it requires ground truth clustering labels, which may not
be available in practice since clustering in generally is used for unsupervised learning.
As input to ``forward`` and ``update`` the metric accepts the following input:
- ``preds`` (:class:`~torch.Tensor`): single integer tensor with shape ``(N,)`` with predicted cluster labels
- ``target`` (:class:`~torch.Tensor`): single integer tensor with shape ``(N,)`` with ground truth cluster labels
As output of ``forward`` and ``compute`` the metric returns the following output:
- ``rand_score`` (:class:`~torch.Tensor`): A tensor with the Rand Score
Args:
kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info.
Example:
>>> import torch
>>> from torchmetrics.clustering import HomogeneityScore
>>> preds = torch.tensor([2, 1, 0, 1, 0])
>>> target = torch.tensor([0, 2, 1, 1, 0])
>>> metric = HomogeneityScore()
>>> metric(preds, target)
tensor(0.4744)
"""
is_differentiable: bool = True
higher_is_better: bool = True
full_state_update: bool = False
plot_lower_bound: float = 0.0
plot_upper_bound: float = 1.0
preds: List[Tensor]
target: List[Tensor]
def __init__(self, **kwargs: Any) -> None:
super().__init__(**kwargs)
self.add_state("preds", default=[], dist_reduce_fx="cat")
self.add_state("target", default=[], dist_reduce_fx="cat")
def update(self, preds: Tensor, target: Tensor) -> None:
"""Update state with predictions and targets."""
self.preds.append(preds)
self.target.append(target)
def compute(self) -> Tensor:
"""Compute rand score over state."""
return homogeneity_score(dim_zero_cat(self.preds), dim_zero_cat(self.target))
def plot(self, val: Union[Tensor, Sequence[Tensor], None] = None, ax: Optional[_AX_TYPE] = None) -> _PLOT_OUT_TYPE:
"""Plot a single or multiple values from the metric.
Args:
val: Either a single result from calling `metric.forward` or `metric.compute` or a list of these results.
If no value is provided, will automatically call `metric.compute` and plot that result.
ax: An matplotlib axis object. If provided will add plot to that axis
Returns:
Figure and Axes object
Raises:
ModuleNotFoundError:
If `matplotlib` is not installed
.. plot::
:scale: 75
>>> # Example plotting a single value
>>> import torch
>>> from torchmetrics.clustering import HomogeneityScore
>>> metric = HomogeneityScore()
>>> metric.update(torch.randint(0, 4, (10,)), torch.randint(0, 4, (10,)))
>>> fig_, ax_ = metric.plot(metric.compute())
.. plot::
:scale: 75
>>> # Example plotting multiple values
>>> import torch
>>> from torchmetrics.clustering import HomogeneityScore
>>> metric = HomogeneityScore()
>>> values = [ ]
>>> for _ in range(10):
... values.append(metric(torch.randint(0, 4, (10,)), torch.randint(0, 4, (10,))))
>>> fig_, ax_ = metric.plot(values)
"""
return self._plot(val, ax)
class CompletenessScore(Metric):
r"""Compute `Completeness Score`_.
A clustering result satisfies completeness if all the data points that are members of a given class are elements of
the same cluster. The metric is not symmetric, therefore swapping ``preds`` and ``target`` yields a different score.
This clustering metric is an extrinsic measure, because it requires ground truth clustering labels, which may not
be available in practice since clustering in generally is used for unsupervised learning.
As input to ``forward`` and ``update`` the metric accepts the following input:
- ``preds`` (:class:`~torch.Tensor`): single integer tensor with shape ``(N,)`` with predicted cluster labels
- ``target`` (:class:`~torch.Tensor`): single integer tensor with shape ``(N,)`` with ground truth cluster labels
As output of ``forward`` and ``compute`` the metric returns the following output:
- ``rand_score`` (:class:`~torch.Tensor`): A tensor with the Rand Score
Args:
kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info.
Example:
>>> import torch
>>> from torchmetrics.clustering import CompletenessScore
>>> preds = torch.tensor([2, 1, 0, 1, 0])
>>> target = torch.tensor([0, 2, 1, 1, 0])
>>> metric = CompletenessScore()
>>> metric(preds, target)
tensor(0.4744)
"""
is_differentiable: bool = True
higher_is_better: bool = True
full_state_update: bool = False
plot_lower_bound: float = 0.0
plot_upper_bound: float = 1.0
preds: List[Tensor]
target: List[Tensor]
def __init__(self, **kwargs: Any) -> None:
super().__init__(**kwargs)
self.add_state("preds", default=[], dist_reduce_fx="cat")
self.add_state("target", default=[], dist_reduce_fx="cat")
def update(self, preds: Tensor, target: Tensor) -> None:
"""Update state with predictions and targets."""
self.preds.append(preds)
self.target.append(target)
def compute(self) -> Tensor:
"""Compute rand score over state."""
return completeness_score(dim_zero_cat(self.preds), dim_zero_cat(self.target))
def plot(self, val: Union[Tensor, Sequence[Tensor], None] = None, ax: Optional[_AX_TYPE] = None) -> _PLOT_OUT_TYPE:
"""Plot a single or multiple values from the metric.
Args:
val: Either a single result from calling `metric.forward` or `metric.compute` or a list of these results.
If no value is provided, will automatically call `metric.compute` and plot that result.
ax: An matplotlib axis object. If provided will add plot to that axis
Returns:
Figure and Axes object
Raises:
ModuleNotFoundError:
If `matplotlib` is not installed
.. plot::
:scale: 75
>>> # Example plotting a single value
>>> import torch
>>> from torchmetrics.clustering import CompletenessScore
>>> metric = CompletenessScore()
>>> metric.update(torch.randint(0, 4, (10,)), torch.randint(0, 4, (10,)))
>>> fig_, ax_ = metric.plot(metric.compute())
.. plot::
:scale: 75
>>> # Example plotting multiple values
>>> import torch
>>> from torchmetrics.clustering import CompletenessScore
>>> metric = CompletenessScore()
>>> values = [ ]
>>> for _ in range(10):
... values.append(metric(torch.randint(0, 4, (10,)), torch.randint(0, 4, (10,))))
>>> fig_, ax_ = metric.plot(values)
"""
return self._plot(val, ax)
class VMeasureScore(Metric):
r"""Compute `V-Measure Score`_.
The V-measure is the harmonic mean between homogeneity and completeness:
..math::
v = \frac{(1 + \beta) * homogeneity * completeness}{\beta * homogeneity + completeness}
where :math:`\beta` is a weight parameter that defines the weight of homogeneity in the harmonic mean, with the
default value :math:`\beta=1`. The V-measure is symmetric, which means that swapping ``preds`` and ``target`` does
not change the score.
This clustering metric is an extrinsic measure, because it requires ground truth clustering labels, which may not
be available in practice since clustering in generally is used for unsupervised learning.
As input to ``forward`` and ``update`` the metric accepts the following input:
- ``preds`` (:class:`~torch.Tensor`): single integer tensor with shape ``(N,)`` with predicted cluster labels
- ``target`` (:class:`~torch.Tensor`): single integer tensor with shape ``(N,)`` with ground truth cluster labels
As output of ``forward`` and ``compute`` the metric returns the following output:
- ``rand_score`` (:class:`~torch.Tensor`): A tensor with the Rand Score
Args:
beta: Weight parameter that defines the weight of homogeneity in the harmonic mean
kwargs: Additional keyword arguments, see :ref:`Metric kwargs` for more info.
Example::
>>> import torch
>>> from torchmetrics.clustering import VMeasureScore
>>> preds = torch.tensor([2, 1, 0, 1, 0])
>>> target = torch.tensor([0, 2, 1, 1, 0])
>>> metric = VMeasureScore(beta=2.0)
>>> metric(preds, target)
tensor(0.4744)
"""
is_differentiable: bool = True
higher_is_better: bool = True
full_state_update: bool = False
plot_lower_bound: float = 0.0
plot_upper_bound: float = 1.0
preds: List[Tensor]
target: List[Tensor]
def __init__(self, beta: float = 1.0, **kwargs: Any) -> None:
super().__init__(**kwargs)
if not (isinstance(beta, float) and beta > 0):
raise ValueError(f"Argument `beta` should be a positive float. Got {beta}.")
self.beta = beta
self.add_state("preds", default=[], dist_reduce_fx="cat")
self.add_state("target", default=[], dist_reduce_fx="cat")
def update(self, preds: Tensor, target: Tensor) -> None:
"""Update state with predictions and targets."""
self.preds.append(preds)
self.target.append(target)
def compute(self) -> Tensor:
"""Compute rand score over state."""
return v_measure_score(dim_zero_cat(self.preds), dim_zero_cat(self.target), beta=self.beta)
def plot(self, val: Union[Tensor, Sequence[Tensor], None] = None, ax: Optional[_AX_TYPE] = None) -> _PLOT_OUT_TYPE:
"""Plot a single or multiple values from the metric.
Args:
val: Either a single result from calling `metric.forward` or `metric.compute` or a list of these results.
If no value is provided, will automatically call `metric.compute` and plot that result.
ax: An matplotlib axis object. If provided will add plot to that axis
Returns:
Figure and Axes object
Raises:
ModuleNotFoundError:
If `matplotlib` is not installed
.. plot::
:scale: 75
>>> # Example plotting a single value
>>> import torch
>>> from torchmetrics.clustering import VMeasureScore
>>> metric = VMeasureScore()
>>> metric.update(torch.randint(0, 4, (10,)), torch.randint(0, 4, (10,)))
>>> fig_, ax_ = metric.plot(metric.compute())
.. plot::
:scale: 75
>>> # Example plotting multiple values
>>> import torch
>>> from torchmetrics.clustering import VMeasureScore
>>> metric = VMeasureScore()
>>> values = [ ]
>>> for _ in range(10):
... values.append(metric(torch.randint(0, 4, (10,)), torch.randint(0, 4, (10,))))
>>> fig_, ax_ = metric.plot(values)
"""
return self._plot(val, ax)