tft.tukey_scale

Computes the scale of the values of a Tensor over the whole dataset.

tft.tukey_scale(
    x: common_types.TensorType,
    reduce_instance_dims: Optional[bool] = True,
    output_dtype: Optional[tf.DType] = None,
    name: Optional[str] = None
) -> tf.Tensor

This computes the scale of x, assuming a Tukey HH distribution, i.e. (x - tukey_location) / tukey_scale is a Tukey HH distribution with parameters tukey_h_params. See the following publication for the definition of the Tukey HH distribution:

Todd C. Headrick, and Mohan D. Pant. "Characterizing Tukey h and hh-Distributions through L-Moments and the L-Correlation," ISRN Applied Mathematics, vol. 2012, 2012. doi:10.5402/2012/980153

x A Tensor, SparseTensor, or RaggedTensor. Its type must be floating point (float{16|32|64}), or integral ([u]int{8|16|32|64}).
reduce_instance_dims By default collapses the batch and instance dimensions to arrive at a single scalar output. If False, only collapses the batch dimension and outputs a vector of the same shape as the input.
output_dtype (Optional) If not None, casts the output tensor to this type.
name (Optional) A name for this operation.

A Tensor containing the scale. If x is floating point, the location will have the same type as x. If x is integral, the output is cast to float32.

TypeError If the type of x is not supported.