Accumulate accuracy statistics. For example, if labels is [1, 2, 3, 4] and predictions is [0, 2, 3, 4] then the accuracy is 3/4 or .75. If the weights were specified as [1, 1, 0, 0] then the accuracy would be 1/2 or .5. `labels` and `predictions` should have the same shape and typ
(self, labels, predictions, weights=None)
| 374 | super(Accuracy, self).__init__(name=name, dtype=dtype) |
| 375 | |
| 376 | def call(self, labels, predictions, weights=None): |
| 377 | """Accumulate accuracy statistics. |
| 378 | |
| 379 | For example, if labels is [1, 2, 3, 4] and predictions is [0, 2, 3, 4] |
| 380 | then the accuracy is 3/4 or .75. If the weights were specified as |
| 381 | [1, 1, 0, 0] then the accuracy would be 1/2 or .5. |
| 382 | |
| 383 | `labels` and `predictions` should have the same shape and type. |
| 384 | |
| 385 | Args: |
| 386 | labels: Tensor with the true labels for each example. One example |
| 387 | per element of the Tensor. |
| 388 | predictions: Tensor with the predicted label for each example. |
| 389 | weights: Optional weighting of each example. Defaults to 1. |
| 390 | |
| 391 | Returns: |
| 392 | The arguments, for easy chaining. |
| 393 | """ |
| 394 | check_ops.assert_equal( |
| 395 | array_ops.shape(labels), array_ops.shape(predictions), |
| 396 | message="Shapes of labels and predictions are unequal") |
| 397 | matches = math_ops.equal(labels, predictions) |
| 398 | matches = math_ops.cast(matches, self.dtype) |
| 399 | super(Accuracy, self).call(matches, weights=weights) |
| 400 | if weights is None: |
| 401 | return labels, predictions |
| 402 | return labels, predictions, weights |
| 403 | |
| 404 | |
| 405 | class CategoricalAccuracy(Mean): |
nothing calls this directly
no test coverage detected