Ficus provides a context manager for matplotlib figures. It finds particular utility in jupyter notebooks, where the user might prefer that figures are not automatically displayed but still wants the option to embed them inline -- such behavior can be accomplished with:
%config InlineBackend.close_figures = False
However, the user is now responsible for making new Figure objects and closing them, which is a pain. The ficus context manager takes care of this boilerplate of opening, showing, saving, and closing figures for you.
Some very basic usage:
from ficus import FigureManager
import numpy as np
X = np.arange(0, 10, .1)
Y = np.exp(X)
with FigureManager(filename='myplot.png', show=True) as (fig, ax):
ax.plot(X, Y)
ax.set_title('The Exponential Function')
FigureManager uses the pyplot.subplots(..) to generate its axes. Thus, you can specify rows and columns and get an array of Axes objects:
with FigureManager(show=True, nrows=3, ncols=4) as (fig, ax_mat):
for i, row in enumerate(ax_mat):
for j, ax in enumerate(row):
ax.plot(X, X**(i+j))
ax.set_title(r'$y = x^{0}$'.format(i+j), fontsize=14)
In fact, you can pass any keyword arguments you'd like to subplots:
with FigureManager(filename='myplot.png', show=True, figsize=(12,8)) as (fig, ax):
ax.plot(X, Y)
Take a look at the examples to see it being used in a jupyter notebook.