tf.raw_ops.IsotonicRegression

Solves a batch of isotonic regression problems.

tf.raw_ops.IsotonicRegression(
    input,
    output_dtype=tf.dtypes.float32,
    name=None
)

input A Tensor. Must be one of the following types: float32, float64, int32, uint8, int16, int8, int64, bfloat16, uint16, half, uint32, uint64. A (batch_size, dim)-tensor holding a batch of inputs.
output_dtype An optional tf.DType from: tf.half, tf.bfloat16, tf.float32, tf.float64. Defaults to tf.float32. Dtype of the output tensor.

Note on supported input-output type combinations:

  • For floating-point types, the output has the same dtype as the input.
  • For 8-bit and 16-bit integer inputs, the output is a 32-bit float.
  • For 32-bit and 64-bit integer inputs, the output is a 64-bit float.

Using unsupported dtype pairs (for example, input=float64 with output=float32) will result in a "Could not find device for node" error.

name A name for the operation (optional).

A tuple of Tensor objects (output, segments).
output A Tensor of type output_dtype.
segments A Tensor of type int32.