pydiodon.plot_eig

pydiodon.plot_eig(L, k=- 1, frac=True, cum=False, Log=False, dot_size=- 1, col='red', title=None, pr=False, x11=True, plotfile=None, fmt='png')

Plots the singular or eigenvalues

Parameters

L1D numpy array

the eigenvalues of A’A

kinteger

the number of eigenvalues to plot or print

default value is -1, which means all positive eigenvalues

fracboolean

if True, the fraction of the norm of A’A is displayed per eigenvalue

cumboolean

if true, the cumulated eigenvalues are displayed can be combined with frac=True

LogBoolean

if True, the Log of the genvalues is displayed

dot_sizeinteger

if > 0, a dot is displayed per eigenvalue

nb_valinteger

number of eigen,values to display;

if nb_val=-1, all eigenvalues are displayed

colstring

color for the plot

titlestring

title on top of the plot;

if None, no title is displayed

prboolean

prints on the screen the eigenvalues if True

inherits the choices for k, frac and cum.

x11Boolean

if True, the plot is displayed on screen

plotfilestring

if not None, the save is save in a file named plotfile

fmtstring

the format of the plot;

accepted formats are png, eps, pdf

Notes

• The eigenvalues \(L_i\) are such that \(\sum_i L_i = ||A'A||^2\). Then, the quality of representation in axis i is \(L_i/\sum_j L_j\). This is displayed by setting frac=True.

• the cumulated eigenvalues are simply given by \(\psi_i = \sum_{j \leq i}L_j\)

af, 25/10/2017 - revised 19.09.18, 22.10.14