Smoother

wannierberri.smoother.get_smoother(energy: ndarray, smear: float, mode: str | None = None)[source]

Return a smoother that applies for the given energy range. The smoother can be used a function that applies to an array. The axis of the array to be smoothed can be controlled with an argument axis, whose default is 0.

If you calculated some quantity data over a range of efermi and want to apply the Fermi-Dirac smoother at 300 K, run the following code:

smoother = wannierberri.get_smoother(efermi, 300, "Fermi-Dirac")
data_smooth = smoother(data)
Parameters:

  • energy (1D array of float) – The energies on which the data are calculated at. Must be evenly spaced.

  • smear (float) –

    • mode == None: not used.

    • mode == "Fermi-Dirac": Smearing parameter in Kelvin units.

    • mode == "Gaussian": Smearing parameter in eV units.

  • mode (str or None) – Smoother mode. Default: None. Avaliable options: None, "Fermi-Dirac", and "Gaussian".

class wannierberri.smoother.FermiDiracSmoother(E: ndarray, T_Kelvin: float, maxdE: int | float = 8)[source]

Bases: AbstractSmoother

Smoother that uses the derivative of Fermi-Dirac function.

Parameters:

  • E (1D array) – The energies on which the data are calculated at.

  • T_Kelvin (float) – Temperature in Kelvin. Transformed into self.smear, which is in eV.

  • maxdE (flaot) – Determines the width of the convoluting function as (-T * maxdE, T * maxdE)

class wannierberri.smoother.GaussianSmoother(E: ndarray, smear: float, maxdE: int | float = 8)[source]

Bases: AbstractSmoother

Smoother that uses Gaussian function.

Parameters:

  • E (1D array) – The energies on which the data are calculated at.

  • smear (float) – Smearing parameter in eV.

  • maxdE (int) – Determines the width of the convoluting function as (-smear * maxdE, smear * maxdE)

class wannierberri.smoother.VoidSmoother[source]

Bases: AbstractSmoother

Void smoother. When called, do nothing and return the original array.

class wannierberri.smoother.AbstractSmoother(E: ndarray, smear: int | float, maxdE: int | float)[source]

Bases: ABC

Smoother for smoothing an array by convolution. This is an abstract class which cannot by instantiated. Only the specific child classes can be instantiated. Each child class should implement its own version of _params, __init__, and _broaden.

Parameters:

  • E (1D array) – The energies on which the data are calculated at.

  • smear (float) – Smearing parameter in eV.

  • maxdE (int) – Determines the width of the convoluting function as (-smear * maxdE, smear * maxdE).

__call__(A, axis=0)[source]

Apply smoother to A along the given axis

abstract __init__(E: ndarray, smear: int | float, maxdE: int | float)[source]

initialize Smoother parameters

abstract _broaden(E)[source]

The broadening method to be used.

abstract property _params

list of parameters that uniquely define the smoother.