• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
2#
3# Licensed under the Apache License, Version 2.0 (the "License");
4# you may not use this file except in compliance with the License.
5# You may obtain a copy of the License at
6#
7#     http://www.apache.org/licenses/LICENSE-2.0
8#
9# Unless required by applicable law or agreed to in writing, software
10# distributed under the License is distributed on an "AS IS" BASIS,
11# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12# See the License for the specific language governing permissions and
13# limitations under the License.
14# ==============================================================================
15"""Confusion matrix related utilities."""
16
17from __future__ import absolute_import
18from __future__ import division
19from __future__ import print_function
20
21from tensorflow.python.framework import dtypes
22from tensorflow.python.framework import ops
23from tensorflow.python.framework import sparse_tensor
24from tensorflow.python.ops import array_ops
25from tensorflow.python.ops import check_ops
26from tensorflow.python.ops import control_flow_ops
27from tensorflow.python.ops import math_ops
28from tensorflow.python.ops import sparse_ops
29from tensorflow.python.util import deprecation
30from tensorflow.python.util.tf_export import tf_export
31
32
33def remove_squeezable_dimensions(
34    labels, predictions, expected_rank_diff=0, name=None):
35  """Squeeze last dim if ranks differ from expected by exactly 1.
36
37  In the common case where we expect shapes to match, `expected_rank_diff`
38  defaults to 0, and we squeeze the last dimension of the larger rank if they
39  differ by 1.
40
41  But, for example, if `labels` contains class IDs and `predictions` contains 1
42  probability per class, we expect `predictions` to have 1 more dimension than
43  `labels`, so `expected_rank_diff` would be 1. In this case, we'd squeeze
44  `labels` if `rank(predictions) - rank(labels) == 0`, and
45  `predictions` if `rank(predictions) - rank(labels) == 2`.
46
47  This will use static shape if available. Otherwise, it will add graph
48  operations, which could result in a performance hit.
49
50  Args:
51    labels: Label values, a `Tensor` whose dimensions match `predictions`.
52    predictions: Predicted values, a `Tensor` of arbitrary dimensions.
53    expected_rank_diff: Expected result of `rank(predictions) - rank(labels)`.
54    name: Name of the op.
55
56  Returns:
57    Tuple of `labels` and `predictions`, possibly with last dim squeezed.
58  """
59  with ops.name_scope(name, 'remove_squeezable_dimensions',
60                      [labels, predictions]):
61    predictions = ops.convert_to_tensor(predictions)
62    labels = ops.convert_to_tensor(labels)
63    predictions_shape = predictions.get_shape()
64    predictions_rank = predictions_shape.ndims
65    labels_shape = labels.get_shape()
66    labels_rank = labels_shape.ndims
67    if (labels_rank is not None) and (predictions_rank is not None):
68      # Use static rank.
69      rank_diff = predictions_rank - labels_rank
70      if (rank_diff == expected_rank_diff + 1 and
71          predictions_shape.dims[-1].is_compatible_with(1)):
72        predictions = array_ops.squeeze(predictions, [-1])
73      elif (rank_diff == expected_rank_diff - 1 and
74            labels_shape.dims[-1].is_compatible_with(1)):
75        labels = array_ops.squeeze(labels, [-1])
76      return labels, predictions
77
78    # Use dynamic rank.
79    rank_diff = array_ops.rank(predictions) - array_ops.rank(labels)
80    if (predictions_rank is None) or (
81        predictions_shape.dims[-1].is_compatible_with(1)):
82      predictions = control_flow_ops.cond(
83          math_ops.equal(expected_rank_diff + 1, rank_diff),
84          lambda: array_ops.squeeze(predictions, [-1]),
85          lambda: predictions)
86    if (labels_rank is None) or (
87        labels_shape.dims[-1].is_compatible_with(1)):
88      labels = control_flow_ops.cond(
89          math_ops.equal(expected_rank_diff - 1, rank_diff),
90          lambda: array_ops.squeeze(labels, [-1]),
91          lambda: labels)
92    return labels, predictions
93
94
95@tf_export('math.confusion_matrix', v1=[])
96def confusion_matrix(labels,
97                     predictions,
98                     num_classes=None,
99                     weights=None,
100                     dtype=dtypes.int32,
101                     name=None):
102  """Computes the confusion matrix from predictions and labels.
103
104  The matrix columns represent the prediction labels and the rows represent the
105  real labels. The confusion matrix is always a 2-D array of shape `[n, n]`,
106  where `n` is the number of valid labels for a given classification task. Both
107  prediction and labels must be 1-D arrays of the same shape in order for this
108  function to work.
109
110  If `num_classes` is `None`, then `num_classes` will be set to one plus the
111  maximum value in either predictions or labels. Class labels are expected to
112  start at 0. For example, if `num_classes` is 3, then the possible labels
113  would be `[0, 1, 2]`.
114
115  If `weights` is not `None`, then each prediction contributes its
116  corresponding weight to the total value of the confusion matrix cell.
117
118  For example:
119
120  ```python
121    tf.math.confusion_matrix([1, 2, 4], [2, 2, 4]) ==>
122        [[0 0 0 0 0]
123         [0 0 1 0 0]
124         [0 0 1 0 0]
125         [0 0 0 0 0]
126         [0 0 0 0 1]]
127  ```
128
129  Note that the possible labels are assumed to be `[0, 1, 2, 3, 4]`,
130  resulting in a 5x5 confusion matrix.
131
132  Args:
133    labels: 1-D `Tensor` of real labels for the classification task.
134    predictions: 1-D `Tensor` of predictions for a given classification.
135    num_classes: The possible number of labels the classification task can
136                 have. If this value is not provided, it will be calculated
137                 using both predictions and labels array.
138    weights: An optional `Tensor` whose shape matches `predictions`.
139    dtype: Data type of the confusion matrix.
140    name: Scope name.
141
142  Returns:
143    A `Tensor` of type `dtype` with shape `[n, n]` representing the confusion
144    matrix, where `n` is the number of possible labels in the classification
145    task.
146
147  Raises:
148    ValueError: If both predictions and labels are not 1-D vectors and have
149      mismatched shapes, or if `weights` is not `None` and its shape doesn't
150      match `predictions`.
151  """
152  with ops.name_scope(name, 'confusion_matrix',
153                      (predictions, labels, num_classes, weights)) as name:
154    labels, predictions = remove_squeezable_dimensions(
155        ops.convert_to_tensor(labels, name='labels'),
156        ops.convert_to_tensor(
157            predictions, name='predictions'))
158    predictions = math_ops.cast(predictions, dtypes.int64)
159    labels = math_ops.cast(labels, dtypes.int64)
160
161    # Sanity checks - underflow or overflow can cause memory corruption.
162    labels = control_flow_ops.with_dependencies(
163        [check_ops.assert_non_negative(
164            labels, message='`labels` contains negative values')],
165        labels)
166    predictions = control_flow_ops.with_dependencies(
167        [check_ops.assert_non_negative(
168            predictions, message='`predictions` contains negative values')],
169        predictions)
170
171    if num_classes is None:
172      num_classes = math_ops.maximum(math_ops.reduce_max(predictions),
173                                     math_ops.reduce_max(labels)) + 1
174    else:
175      num_classes_int64 = math_ops.cast(num_classes, dtypes.int64)
176      labels = control_flow_ops.with_dependencies(
177          [check_ops.assert_less(
178              labels, num_classes_int64, message='`labels` out of bound')],
179          labels)
180      predictions = control_flow_ops.with_dependencies(
181          [check_ops.assert_less(
182              predictions, num_classes_int64,
183              message='`predictions` out of bound')],
184          predictions)
185
186    if weights is not None:
187      weights = ops.convert_to_tensor(weights, name='weights')
188      predictions.get_shape().assert_is_compatible_with(weights.get_shape())
189      weights = math_ops.cast(weights, dtype)
190
191    shape = array_ops.stack([num_classes, num_classes])
192    indices = array_ops.stack([labels, predictions], axis=1)
193    values = (array_ops.ones_like(predictions, dtype)
194              if weights is None else weights)
195    cm_sparse = sparse_tensor.SparseTensor(
196        indices=indices,
197        values=values,
198        dense_shape=math_ops.cast(shape, dtypes.int64))
199    zero_matrix = array_ops.zeros(math_ops.cast(shape, dtypes.int32), dtype)
200
201    return sparse_ops.sparse_add(zero_matrix, cm_sparse)
202
203
204@tf_export(v1=['math.confusion_matrix', 'confusion_matrix'])
205@deprecation.deprecated_endpoints('confusion_matrix', 'train.confusion_matrix')
206def confusion_matrix_v1(labels,
207                        predictions,
208                        num_classes=None,
209                        dtype=dtypes.int32,
210                        name=None,
211                        weights=None):
212  """Computes the confusion matrix from predictions and labels.
213
214  The matrix columns represent the prediction labels and the rows represent the
215  real labels. The confusion matrix is always a 2-D array of shape `[n, n]`,
216  where `n` is the number of valid labels for a given classification task. Both
217  prediction and labels must be 1-D arrays of the same shape in order for this
218  function to work.
219
220  If `num_classes` is `None`, then `num_classes` will be set to one plus the
221  maximum value in either predictions or labels. Class labels are expected to
222  start at 0. For example, if `num_classes` is 3, then the possible labels
223  would be `[0, 1, 2]`.
224
225  If `weights` is not `None`, then each prediction contributes its
226  corresponding weight to the total value of the confusion matrix cell.
227
228  For example:
229
230  ```python
231    tf.math.confusion_matrix([1, 2, 4], [2, 2, 4]) ==>
232        [[0 0 0 0 0]
233         [0 0 1 0 0]
234         [0 0 1 0 0]
235         [0 0 0 0 0]
236         [0 0 0 0 1]]
237  ```
238
239  Note that the possible labels are assumed to be `[0, 1, 2, 3, 4]`,
240  resulting in a 5x5 confusion matrix.
241
242  Args:
243    labels: 1-D `Tensor` of real labels for the classification task.
244    predictions: 1-D `Tensor` of predictions for a given classification.
245    num_classes: The possible number of labels the classification task can have.
246      If this value is not provided, it will be calculated using both
247      predictions and labels array.
248    dtype: Data type of the confusion matrix.
249    name: Scope name.
250    weights: An optional `Tensor` whose shape matches `predictions`.
251
252  Returns:
253    A `Tensor` of type `dtype` with shape `[n, n]` representing the confusion
254    matrix, where `n` is the number of possible labels in the classification
255    task.
256
257  Raises:
258    ValueError: If both predictions and labels are not 1-D vectors and have
259      mismatched shapes, or if `weights` is not `None` and its shape doesn't
260      match `predictions`.
261  """
262  return confusion_matrix(labels, predictions, num_classes, weights, dtype,
263                          name)
264