Fico

Fico (from Latin ficus, "fig") assists in creating consistent Matplotlib figures with easy scaling and styling groups of figures.

Overview

Figures are organized into several substructures that are combined to form a figure tree:

1. The highest level is the FigureContainer - a figure tree will have one of these at the root.
2. The FigureContainer consists of several FigureCollections - these will always be direct descendants of a FigureContainer.
3. Each FigureCollection contains different subclasses of Figure - likewise, these are always direct descendants of a FigureCollection.

The top-level FigureContainer defines global styling and handles top-level building. All child Figures inherit the styles from the parent container, unless specifically overwritten in the figure-definition. Each instance of FigureCollection corresponds to a single folder in the final build and thus their main purpose is to group a related collection of figures. Each instance of Figure corresponds to a single diagram, and thus a single file when built.

The default export format for figures is pdf.

Running a build

You can run a build locally by running the following script in a terminal

fico build

this will rebuild all figures in publication quality. If instead you want a faster build, you can use draft mode. This will only build those figures initialized with @collection.plot_figure(..., only_build_this=True). Keep in mind that if no figure has this flag set, draft mode is completely equivalent to the standard build mode, and all figures will be built.

fico build -d

Specifying which container to build

By default, fico build builds the container defined in the module index exported as container. This can be overwritten such that no particular project-structure is enforced. Say for example, you have defined a root container in a variable called root in a package-structure like package_name/module_name. This container is then built by specifying the location of the container to the build command:

fico build --container package_name.module_name:root

Note: The path of invokation of Fico is added to sys.path, so the container path should be relative to this point.

Defining figures

The top-level FigureContainer

As shown above, a figure container is passed to the Fico build command, and acts as the root of the figure tree. As such, you'll always need to define at least one FigureContainer to build figures.

from fico import FigureContainer

### Code that instantiates/imports FigureCollections ###

collection = FigureContainer([col1, col2, ...])

Overwriting the default figure builder

A figure builder traverses the entire figure tree, applies global styling to the figures, and outputs them in a directory-structure that directly maps to the structure of the figure tree. A default builder is defined in Fico that is used if it isn't explicitly overwritten when defining the FigureContainer.

# Using SciencePlots to style figures
from fico import FigureContainer, FigureBuilder
import scienceplots

### Code that instantiates/imports FigureCollections ###

collection = FigureContainer([col1, col2, ...], builder=FigureBuilder(style="ieee"))

The grouping FigureCollection

A figure collection just needs a name.

from fico import FigureCollection

collection = FigureCollection("my_collection")

The figures themselves Figure

When you've defined an appropriate FigureCollection, you can start registering figure functions in the collection

import matplotlib.pyplot as plt
import numpy as np


@collection.plot_figure()
def figure1():
    # Some example data to display
    x = np.linspace(0, 2 * np.pi, 400)
    y = np.sin(x**2)

    fig, ax = plt.subplots()
    ax.plot(x, y)
    ax.set_title("A single plot")

    return fig

Scaling figures

By default, the width of a figure is equal to the width of an A4 page with 2.54 cm margins (453pt). The height is automatically calculated such that the ratio between the width and height is the golden ratio. The default settings, equivalent to omitting both the height and width settings are

@collection.plot_figure(width=1.0, height=1.0)
def figure1(fig, ax):

There are a couple options when you want to scale a figure. Both width and height can be specified as a float, in which case the size is relative to the default size. Note that the height scaling is always applied after the figure height has been calculated to conform to the golden ratio, i.e. when height=1.0 the ratio between the width and the height is the golden ratio.

For example, setting the width of a figure to half of a standard page (omitting the height setting, such that it is automatically calculated)

@collection.plot_figure(width=0.5)
def figure1(fig, ax):

Setting the width of a figure to half of a standard page, and cutting the height in half

@collection.plot_figure(width=0.5, height=0.5)
def figure1(fig, ax):

If the relative sizing and the automatic calculation are prohibitive to what you're trying to achieve, both dimensions can be specified absolutely in a string. The supported units are mm, cm, pt and in:

@collection.plot_figure(width="3cm", height="30mm")
def figure1(fig, ax):