Python Sparse data Analysis Package

Note

This page is a reference documentation. It only explains the class signature, and not how to use it. Please refer to the gallery for the big picture.

class pysap.extensions.transform.ISAPWaveletTransformBase(nb_scale, verbose=0, dim=2, padding_mode='zero', **kwargs)[source]¶

Define the structure that will be used to store the ISAP results.

classmethod decimated(nb_scale, iso_shape, nb_band, scale_shift=0)[source]¶

Compute decimated transformation parameters.

Parameters

nb_scale : int

the number of scale of the decomposition that includes the approximation scale.

iso_shape : int

the data isotropic shape.

nb_band : int

the number of band.

scale_shift : int, default 0

decimate the image with a factor of 2**(scale + scale_shift).

Returns

bands_names : list of str

the name of the different bands.

flatten_fct : int

a function used to reorganize the ISAP decomposition coefficients, see ‘pysap/extensions/formating.py’ module for more details.

unflatten_fct : callable

a function used to reorganize the decomposition coefficients using ISAP convention, see ‘pysap/extensions/formating.py’ module for more details.

is_decimated : bool

True if the decomposition include a decimation of the band number of coefficients.

nb_band_per_scale : ndarray (<nb_scale>, )

vector of int holding the number of band per scale.

bands_lengths : ndarray (<nb_scale>, max(<nb_band_per_scale>, 0))

array holding the length between two bands of the data vector per scale.

bands_shapes : list of list of 2-uplet (<nb_scale>, <nb_band_per_scale>)

structure holding the shape of each bands at each scale.

classmethod undecimated(nb_scale, iso_shape, nb_band)[source]¶

Compute undecimated transformation parameters.

Parameters

nb_scale : int

the number of scale of the decomposition that includes the approximation scale.

iso_shape : int

the data isotropic shape.

nb_band : int

the number of band.

Returns

bands_names : list of str

the name of the different bands.

flatten_fct : int

a function used to reorganize the ISAP decomposition coefficients, see ‘pysap/extensions/formating.py’ module for more details.

unflatten_fct : callable

a function used to reorganize the decomposition coefficients using ISAP convention, see ‘pysap/extensions/formating.py’ module for more details.

is_decimated : bool

True if the decomposition include a decimation of the band number of coefficients.

nb_band_per_scale : ndarray (<nb_scale>, )

vector of int holding the number of band per scale.

bands_lengths : ndarray (<nb_scale>, max(<nb_band_per_scale>, 0))

array holding the length between two bands of the data vector per scale.

bands_shapes : list of list of 2-uplet (<nb_scale>, <nb_band_per_scale>)

structure holding the shape of each bands at each scale.