dev module

This module defines the basic stem acquisitions objects used for processing. These objects are:

  1. The 3D spectrum-image,

  2. The 2D HAADF image,

class inpystem.dev.AbstractDev(key, data, mask=None, sigma=None, seed=None, normalize=True, verbose=True)

Abstract Dev acquisition class.

This is an abstract class, which mean you can not instantiate such object.

It defines the structure for a Dev acquisition object.

key: str

1-word description of the Dev2D image.

data: (m,n) or (m, n, l) numpy array

The Dev2D image data before the noise step. Its dimension is (m,n).

ndata: (m,n) or (m, n, l) numpy array

The noised Dev2D image. If snr is None, ndata is None. Its dimension is (m,n).

sigma: float

The noise standard deviation.

seed: optional, int

The random noise matrix seed.

normalize: bool

If :code:normalize` is True, the data will be centered and normalize before the corruption steps.

mean_std: None, 2-tuple

It stores the data mean and std in case normalize is True.

verbose: bool

If True, information will be displayed. Default is True.

__init__(key, data, mask=None, sigma=None, seed=None, normalize=True, verbose=True)

AbstractDev constructor.

Parameters

  • key (str) – 1-word description of the Dev2D image. Generally, it’s common to the stem acquisition object.

  • data ((m, n) or (m, n, l) numpy array) – The noise-free image data.

  • mask ((m, n) numpy array) – The sampling mask.

  • sigma (optional, None, float) – The desired standard deviation used to model noise. Dafault is None for no additional noise.

  • seed (optional, None, int) – The random noise matrix seed. Dafault is None for no seed initialization.

  • normalize (optional, bool) – If :code:normalize` is True, the data will be centered and normalize before the corruption steps. Default is True.

  • verbose (optional, bool) – If True, information will be displayed. Default is True.

property seed

seed property getter.

property sigma

sigma property getter.

set_ndata()

Constructs the noised data.

It is also used to draw a new noise matrix.

class inpystem.dev.Dev2D(key, hsdata, scan=None, modif_file=None, sigma=None, seed=None, normalize=True, verbose=True)

Dev2D Class.

Variables

  • key (str) – 1-word description of the Dev2D image.

  • hsdata (Signal2D hyperspy data) – The hyperspy Signal2D image. Its dimension is denoted (m,n). This is used to communicate with the parrent class.

  • data ((m,n) numpy array) – The Dev2D image data before the noise step. Its dimension is (m,n).

  • ndata ((m,n) numpy array) – The noised Dev2D image. If snr is None, ndata is None. Its dimension is (m,n).

  • scan (optional, Scan object) – The sampling scan object associated with the data. Default is None for full sampling.

  • sigma (float) – The noise standard deviation.

  • seed (optional, int) – The random noise matrix seed.

  • normalize (bool) – If :code:normalize` is True, the data will be centered and normalize before the corruption steps.

  • mean_std (None, 2-tuple) – It stores the data mean and std in case normalize is True.

  • verbose (bool) – If True, information will be displayed. Default is True.

__init__(key, hsdata, scan=None, modif_file=None, sigma=None, seed=None, normalize=True, verbose=True)

SpectrumImage constructor.

Parameters

  • key (str) – 1-word description of the Dev2D image. Generally, it’s common to the stem acquisition object.

  • hsdata (Signal2D hyperspy data) – The noise-free Dev2D image data. Its dimension is denoted (m,n).

  • scan (optional, None, Scan object) – The sampling scan object associated with the data. Default is None for full sampling.

  • modif_file (optional, None, str) – A .conf configuration file to remove rows, columns or dead pixels. Default is None for no modification.

  • sigma (optional, None, float) – The desired standard deviation used to model noise. Dafault is None for no additional noise.

  • seed (optional, None, int) – The random noise matrix seed. Dafault is None for no seed initialization.

  • normalize (optional, bool) – If :code:normalize` is True, the data will be centered and normalize before the corruption steps. Default is True.

  • verbose (optional, bool) – If True, information will be displayed. Default is True.

restore(method='interpolation', parameters={}, verbose=None)
plot(noised=False)

Plots the haadf image.

Parameters

noised (optional, bool) – If True, the noised data is used. If False, the noise-free data is shown. Default is False.

class inpystem.dev.Dev3D(key, hsdata, scan=None, modif_file=None, sigma=None, seed=None, normalize=True, PCA_transform=False, PCA_th='auto', verbose=True)

Dev3D Class

Variables

  • key (str) – 1-word description of the Dev3D.

  • hsdata (Signal1D hyperspy data) – The hyperspy Signal2D image. Its dimension is denoted (m,n). This is used to communicate with the parrent class.

  • data ((m,n,l) numpy array) – The Dev3D data before the noise step. Its dimension is (m,n,l).

  • ndata ((m,n,l) numpy array) – The noised Dev3D data. If snr is None, ndata is None. Its dimension is (m,n,l).

  • snr (optional, float) – The desired snr used for the noising step.

  • sigma (float) – The noise standard deviation.

  • seed (optional, int) – The random noise matrix seed.

  • normalize (bool) – If normalize is True, the data will be centered and normalize before the corruption steps.

  • PCA_transform (bool) – If PCA_transformed is True, a PCA transformation has been applied to the data.

  • PCA_info (None, dictionary) – If PCA_transformed is True, PCA_info contains informations about the reduction. Otherwise, it is None.

  • PCA_operator (PcaHandler) – The PCA operator.

  • verbose (bool) – If True, information will be displayed. Default is True.

__init__(key, hsdata, scan=None, modif_file=None, sigma=None, seed=None, normalize=True, PCA_transform=False, PCA_th='auto', verbose=True)

Dev3D __init__ function.

Parameters

  • key (str) – 1-word description of the Dev3D image. Generally, it’s common to the stem acquisition object.

  • hsdata (Signal1D hyperspy data) – The noise-free Dev3D image data. Its dimension is denoted (m,n, l).

  • scan (optional, None, Scan object) – The sampling scan object associated with the data. Default is None for full sampling.

  • modif_file (optional, None, str) – A .conf configuration file to remove rows, columns or dead pixels. Default is None for no modification.

  • sigma (optional, None, float) – The desired standard deviation used to model noise. Dafault is None for no additional noise.

  • seed (optional, None, int) – The random noise matrix seed. Dafault is None for no seed initialization.

  • normalize (optional, bool) – If :code:normalize` is True, the data will be centered and normalize before the corruption steps. Default is True.

  • PCA_transform (optional, bool) – If PCA_transformed is True, a PCA transformation is applied to the data. Default is False.

  • PCA_th (optional, str, int) –

    The desired data dimension after dimension reduction. Possible values are:

    • ’auto’ for automatic choice,

    • ’max’ for maximum value

    • an int value for user value.

    Default is ‘auto’.

  • verbose (optional, bool) – If True, information will be displayed. Default is True.

direct_transform(data)

Applies the Dev3D PCA transformation and normalization steps to data.

Parameters

data ((m, n, l) numpy array, hs image) – Data whose shape is the same as self.data.

inverse_transform(data)

Applies the Dev3D PCA inverse transformation and inverse normalization steps to spim.

Parameters

data ((m, n, l) numpy array, hs image) – Data whose shape is the same as self.data.

restore(method='interpolation', parameters={}, PCA_transform=None, PCA_th='auto', verbose=None)
show_sum(noised=False)

Shows the sum of the data along the last axis.

Parameters

noised (optional, bool) – If True, the noised data is used. If False, the noise-free data is shown. Default is False.

plot_as2D(noised=False)

Implements the HypersSpy tool to visualize the image for a given band.

Parameters

noised (optional, bool) – If True, the noised data is used. If False, the noise-free data is shown. Default is False.

plot_as1D(noised=False)

Implements the HypersSpy tool to visualize the spectrum for a given pixel.

Parameters

noised (optional, bool) – If True, the noised data is used. If False, the noise-free data is shown. Default is False.

plot_roi(noised=False)

Implements the Hyperspy tool to analyse regions of interest.

Parameters

noised (optional, bool) – If True, the noised data is used. If False, the noise-free data is shown. Default is False.