Subplots / Tiled plots

There are three ways to create subplot layouts with Plopp:

• Using the in-built tiled functionality

• Using manually created Matplotlib axes

• Placing figures inside container widgets

The first two methods will work with any (static or interactive) backend, while the third will only work with the interactive (%matplotlib widget) backend.

Using tiled:

  New in version 23.08.0.

Leveraging figure operator overloading

The fastest way to make simple subplots is to use the + and / operators on the figures you create.

Adding two figures will place them side-by-side, while dividing a figure by another will stack them vertically.

%matplotlib inline
import plopp as pp
import matplotlib.pyplot as plt

p1 = pp.data.data1d().plot()
p2 = pp.data.data2d().plot()

p1 + p2

p1 / p2

(p1 + p2) / (p2 + p1)

More complex configurations

While the figure operators are using tiled under-the-hood, it is also possible to use tiled directly.

Its use is based on Matplotlib’s GridSpec.

tiled = pp.tiled(nrows=1, ncols=3)
tiled[0, :2] = p1
tiled[0, 2] = p2
tiled

Using Matplotlib axes

%matplotlib inline
import plopp as pp
import matplotlib.pyplot as plt

da = pp.data.data3d()

To create our subplots, we use Matplotlib’s in-built plt.subplots function. We then pass each subplot axes to Plopp’s plot function as the ax argument:

fig, ax = plt.subplots(2, 2, figsize=(8, 6))

p1 = pp.plot(da['z', 0], ax=ax[0, 0])
p2 = pp.plot(da['z', 0]['y', 0], ax=ax[0, 1])
p3 = pp.plot(da['x', 0]['y', 0], ax=ax[1, 0])
p4 = pp.plot(da['x', -1], cmap='magma', ax=ax[1, 1])

# Matplotlib's utility to optimize spacing between subplots
fig.tight_layout()

Using container widgets

When the interactive backend is in use, each figure is a widget that can be placed inside other ipywidgets for controlling the layout. In this case, each subplot will have its own toolbar.

%matplotlib widget
import ipywidgets as ipw

Given four plots

subplot_size = (4, 3)

p1 = pp.plot(da['z', 0], figsize=subplot_size)
p2 = pp.plot(da['z', 0]['y', 0], figsize=subplot_size)
p3 = pp.plot(da['x', 0]['y', 0], figsize=subplot_size)
p4 = pp.plot(da['x', -1], cmap='magma', figsize=subplot_size)

we can create a figure with subplots using

ipw.VBox([ipw.HBox([p1, p2]), ipw.HBox([p3, p4])])

Plopp actually provides a Box widget as a shorthand for combining ipw.VBox and ipw.HBox:

from plopp.widgets import Box

Box([[p1, p2], [p3, p4]])

Warning

When using ipw.VBox or Box, alignment of axes is not ensured. This method for creating subplots is applicable for quickly creating interactive interfaces, but we do not recommend it for publication purposes.

Subplots with 3D scatter plots

The three-dimensional visualizations in Plopp do not use Matplotlib, but Pythreejs. Consequently, it is not possible to insert a three-dimensional scatter plot inside a subplot made of Matplotlib axes. The only way to make a three-dimensional view part of subplots is to use the container widgets described above.

da1 = pp.data.data1d()
da2 = pp.data.scatter()

p1 = pp.plot(da1)
p2 = pp.scatter3d(da2, x='x', y='y', z='z', figsize=(550, 400))

Box([p1, p2])