Python tensorflow.python.ops.nn.top_k() Examples

def sparsemax(logits, name=None):
  """Computes sparsemax activations [1].

  For each batch `i` and class `j` we have
    sparsemax[i, j] = max(logits[i, j] - tau(logits[i, :]), 0)

  [1]: https://arxiv.org/abs/1602.02068

  Args:
    logits: A `Tensor`. Must be one of the following types: `half`, `float32`,
      `float64`.
    name: A name for the operation (optional).

  Returns:
    A `Tensor`. Has the same type as `logits`.
  """

  with ops.name_scope(name, "sparsemax", [logits]) as name:
    logits = ops.convert_to_tensor(logits, name="logits")
    obs = array_ops.shape(logits)[0]
    dims = array_ops.shape(logits)[1]

    z = logits - math_ops.reduce_mean(logits, axis=1)[:, array_ops.newaxis]

    # sort z
    z_sorted, _ = nn.top_k(z, k=dims)

    # calculate k(z)
    z_cumsum = math_ops.cumsum(z_sorted, axis=1)
    k = math_ops.range(
        1, math_ops.cast(dims, logits.dtype) + 1, dtype=logits.dtype)
    z_check = 1 + k * z_sorted > z_cumsum
    # because the z_check vector is always [1,1,...1,0,0,...0] finding the
    # (index + 1) of the last `1` is the same as just summing the number of 1.
    k_z = math_ops.reduce_sum(math_ops.cast(z_check, dtypes.int32), axis=1)

    # calculate tau(z)
    indices = array_ops.stack([math_ops.range(0, obs), k_z - 1], axis=1)
    tau_sum = array_ops.gather_nd(z_cumsum, indices)
    tau_z = (tau_sum - 1) / math_ops.cast(k_z, logits.dtype)

    # calculate p
    return math_ops.maximum(
        math_ops.cast(0, logits.dtype), z - tau_z[:, array_ops.newaxis]) 

def streaming_sparse_average_precision_at_top_k(top_k_predictions,
                                                labels,
                                                weights=None,
                                                metrics_collections=None,
                                                updates_collections=None,
                                                name=None):
  """Computes average precision@k of predictions with respect to sparse labels.

  `streaming_sparse_average_precision_at_top_k` creates two local variables,
  `average_precision_at_<k>/total` and `average_precision_at_<k>/max`, that
  are used to compute the frequency. This frequency is ultimately returned as
  `average_precision_at_<k>`: an idempotent operation that simply divides
  `average_precision_at_<k>/total` by `average_precision_at_<k>/max`.

  For estimation of the metric over a stream of data, the function creates an
  `update_op` operation that updates these variables and returns the
  `precision_at_<k>`. Set operations applied to `top_k` and `labels` calculate
  the true positives and false positives weighted by `weights`. Then `update_op`
  increments `true_positive_at_<k>` and `false_positive_at_<k>` using these
  values.

  If `weights` is `None`, weights default to 1. Use weights of 0 to mask values.

  Args:
    top_k_predictions: Integer `Tensor` with shape [D1, ... DN, k] where N >= 1.
      Commonly, N=1 and `predictions_idx` has shape [batch size, k]. The final
      dimension must be set and contains the top `k` predicted class indices.
      [D1, ... DN] must match `labels`. Values should be in range [0,
      num_classes).
    labels: `int64` `Tensor` or `SparseTensor` with shape [D1, ... DN,
      num_labels] or [D1, ... DN], where the latter implies num_labels=1. N >= 1
      and num_labels is the number of target classes for the associated
      prediction. Commonly, N=1 and `labels` has shape [batch_size, num_labels].
      [D1, ... DN] must match `top_k_predictions`. Values should be in range [0,
      num_classes).
    weights: `Tensor` whose rank is either 0, or n-1, where n is the rank of
      `labels`. If the latter, it must be broadcastable to `labels` (i.e., all
      dimensions must be either `1`, or the same as the corresponding `labels`
      dimension).
    metrics_collections: An optional list of collections that values should be
      added to.
    updates_collections: An optional list of collections that updates should be
      added to.
    name: Name of new update operation, and namespace for other dependent ops.

  Returns:
    mean_average_precision: Scalar `float64` `Tensor` with the mean average
      precision values.
    update: `Operation` that increments variables appropriately, and whose
      value matches `metric`.

  Raises:
    ValueError: if the last dimension of top_k_predictions is not set.
  """
  return metrics_impl._streaming_sparse_average_precision_at_top_k(  # pylint: disable=protected-access
      predictions_idx=top_k_predictions,
      labels=labels,
      weights=weights,
      metrics_collections=metrics_collections,
      updates_collections=updates_collections,
      name=name)