tf.math.sobol_sample

View source on GitHub

Generates points from the Sobol sequence.

tf.math.sobol_sample(
    dim, num_results, skip=0, dtype=tf.dtypes.float32, name=None
)

Creates a Sobol sequence with num_results samples. Each sample has dimension dim. Skips the first skip samples.

dim Positive scalar Tensor representing each sample's dimension.
num_results Positive scalar Tensor of dtype int32. The number of Sobol points to return in the output.
skip (Optional) Positive scalar Tensor of dtype int32. The number of initial points of the Sobol sequence to skip. Default value is 0.
dtype (Optional) The tf.Dtype of the sample. One of: tf.float32 or tf.float64. Defaults to tf.float32.
name (Optional) Python str name prefixed to ops created by this function.

Tensor of samples from Sobol sequence with shape [num_results, dim].