Class ClassificationError

tensorpack/callbacks/inference.py:135–177

Compute **true** classification error in batch mode, from a ``wrong`` tensor. The ``wrong`` tensor is supposed to be an binary vector containing whether each sample in the batch is *incorrectly* classified. You can use ``tf.nn.in_top_k`` to produce this vector. This Inferencer

class ClassificationError(Inferencer):
    """
    Compute **true** classification error in batch mode, from a ``wrong`` tensor.

    The ``wrong`` tensor is supposed to be an binary vector containing
    whether each sample in the batch is *incorrectly* classified.
    You can use ``tf.nn.in_top_k`` to produce this vector.

    This Inferencer produces the "true" error, which could be different from
    ``ScalarStats('error_rate')``.
    It takes account of the fact that batches might not have the same size in
    testing (because the size of test set might not be a multiple of batch size).
    Therefore the result can be different from averaging the error rate of each batch.

    You can also use the "correct prediction" tensor, then this inferencer will
    give you "classification accuracy" instead of error.
    """

    def __init__(self, wrong_tensor_name='incorrect_vector', summary_name='validation_error'):
        """
        Args:
            wrong_tensor_name(str): name of the ``wrong`` binary vector tensor.
            summary_name(str): the name to log the error with.
        """
        self.wrong_tensor_name = wrong_tensor_name
        self.summary_name = summary_name

    def _before_inference(self):
        self.err_stat = RatioCounter()

    def _get_fetches(self):
        return [self.wrong_tensor_name]

    def _on_fetches(self, outputs):
        vec = outputs[0]
        # TODO put shape assertion into inference-runner
        assert vec.ndim == 1, "{} is not a vector!".format(self.wrong_tensor_name)
        batch_size = len(vec)
        wrong = np.sum(vec)
        self.err_stat.feed(wrong, batch_size)

    def _after_inference(self):
        return {self.summary_name: self.err_stat.ratio}


class BinaryClassificationStats(Inferencer):