alibi_detect.utils.tensorflow.misc

Functions

clone_model

clone_model(model: keras.src.models.model.Model) -> keras.src.models.model.Model

Clone a sequential, functional or subclassed tf.keras.Model.

Name	Type	Default	Description
model	keras.src.models.model.Model

Returns

• Type: keras.src.models.model.Model

quantile

quantile(sample: tensorflow.python.framework.tensor.Tensor, p: float, type: int = 7, sorted: bool = False) -> float

Estimate a desired quantile of a univariate distribution from a vector of samples

Name	Type	Default	Description
sample	tensorflow.python.framework.tensor.Tensor		A 1D vector of values
p	float		The desired quantile in (0,1)
type	int	7	The method for computing the quantile. See https://wikipedia.org/wiki/Quantile#Estimating_quantiles_from_a_sample
sorted	bool	False	Whether or not the vector is already sorted into ascending order

Returns

• Type: float

subset_matrix

subset_matrix(mat: tensorflow.python.framework.tensor.Tensor, inds_0: tensorflow.python.framework.tensor.Tensor, inds_1: tensorflow.python.framework.tensor.Tensor) -> tensorflow.python.framework.tensor.Tensor

Take a matrix and return the submatrix correspond to provided row and column indices

Name	Type	Default	Description
mat	tensorflow.python.framework.tensor.Tensor		A 2D matrix
inds_0	tensorflow.python.framework.tensor.Tensor		A vector of row indices
inds_1	tensorflow.python.framework.tensor.Tensor		A vector of column indices

Returns

• Type: tensorflow.python.framework.tensor.Tensor

zero_diag

zero_diag(mat: tensorflow.python.framework.tensor.Tensor) -> tensorflow.python.framework.tensor.Tensor

Set the diagonal of a matrix to 0

Name	Type	Default	Description
mat	tensorflow.python.framework.tensor.Tensor		A 2D square matrix

Returns

• Type: tensorflow.python.framework.tensor.Tensor