Setting themes
by: Kevin Broløs
(Feyn version 3.4.0 or newer)
In some situations, such as if you're running in a dark mode editor, or if you're looking to publish your results in a journal, you might want to use a different theme.
Example
import feyn
from feyn import Theme
ql = feyn.QLattice()
models = ql.sample_models(['Hello'], 'World')
We have four themes available: Default, Dark, Monochrome and Monochrome Dark
Theme.set_theme('default')
models[0]
These themes also apply to our plots and graphs. We've made a comparison to the stylesheet reference in matplotlib down below, and also showcase some colormaps that are registered with matplotlib.
Theme.set_theme('dark')
models[0]
These themes also apply to our plots and graphs. We've made a comparison to the stylesheet reference in matplotlib down below, and also showcase some colormaps that are registered with matplotlib.
Theme.set_theme('mono')
models[0]
These themes also apply to our plots and graphs. We've made a comparison to the stylesheet reference in matplotlib down below, and also showcase some colormaps that are registered with matplotlib.
Theme.set_theme('mono_dark')
models[0]
These themes also apply to our plots and graphs. We've made a comparison to the stylesheet reference in matplotlib down below, and also showcase some colormaps that are registered with matplotlib.
Full list of feyn colormaps: feyn, feyn-diverging, feyn-partial, feyn-primary, feyn-highlight, feyn-secondary, feyn-accent, feyn-signal.
This should help make our graphs a little easier on the eyes in different editors.
Flipping colormaps
Some colormaps, notably feyn-diverging, are used heavily for classification plots where color coding can carry important meaning.
For a binary outcome, sometimes 0 is a good outcome and 1 is bad, and other times, it's the opposite. The default ordering is 0 gets the "danger" hues and 1 gets the "safer" hues.
There's a helper function on Theme to flip the order of this colormap, so you can easily control the positive and negative end of the color scale without having to supply a new colormap for each plot you do throughout.
As a shorthand for the diverging colormap:
Theme.flip_diverging_cmap()
which is the colormap used by default in plot_response_2d and plot_probability_scores.
or as a general function that works for all of the feyn colormaps, should you want that:
Theme.flip_cmap("feyn-diverging")
Each time you call the function, the colormap is reversed in order, so calling it twice will revert it to normal.