pydiodon.plot_components_heatmap

pydiodon.plot_components_heatmap(Y, axis_1=1, axis_2=2, bins=256, cmap='ocean', range=None, log=True, scale=False, title=None, x11=True, imfile=None, fmt='png')

Builds the density heatmap of the point cloud of the components of a PCA

Parameters

Ya \(n \times k\) 2D numpy array

array of the components

binsan integer

the size of the image, in pixels

axis_1an integer

first axis to be plotted

columns in the image

axis_2an integer

second axis of Y

rows in the image

binsan integer

the size of the image

it is also the number of classes per axis in the 2D histogram

cmapa string

the matplotlib colormap for the heatmap

default value is ocean

classical default value in matplotlib is viridis

see https://matplotlib.org/stable/tutorials/colors/colormaps.html for a choice

rangeis None

compulsory for callind 2D histogram

loga boolean

whether to translate the densities in a logarithmic scale

default value is True

scalea boolean

deprecated

default value is False

titlea string

the title of the heatmap

default value is None (no title)

x11a boolean

whether to display the plot on the screen

default value is True (plot displayed)

plotfilea string

the name of the file where to save the plot

default value is None (plot not saved)

fmta string

format of the file where the plot is saved

possible values are : png, eps, pdf

default value is png

Returns

a heatmap

Notes

Y is a numpy array of coordinates, as produced by the MDS

when two axis are selected

• the two components are extracted

• a 2D histogram of their values is computed with a call to np.histogram2d()

• the heatmap is drawn with the values of the 2D histogram

af, 25/01/2020, revised 22.10.36