Creating a figure with plots, sliders and other widgets arranged with a gridspec layout

Question

I want to create an animated graph which parameters can be controlled with sliders and other widgets. I have to create several similar figures so I want to pack this in some class to be reused with various parameters. But before that, I wanted to figure out how it even works.

This code will create a graph on upper part of the figure and will leave the rest blank. However, the x and y axes are drawn in range [-0.05,0.05], instead of in predefined ranges below.

How do I make sure the graph is drawn to scale I want?

Another thing I don't know is how do I add widgets to the layout? I want to insert them into gridspec without hardcoding the coordinates and sizes, to have them adjust to given space.

I tried something below, but that obviously didn't work. How do I go about this to make it work as I want?

import matplotlib.gridspec as gridspec
import numpy as np

from matplotlib import pyplot as plt

PI = np.pi

# Half width of the graph x-axis
x_axis = 4*PI
# x_axis offset
x_offset = 0
# Half height of the graph y-axis
y_axis = 8
# y_axis offset
y_offset = -1

fig = plt.figure()

mainGrid = gridspec.GridSpec(2, 1)
graphCell = plt.subplot(mainGrid[0, :])
graphCell.plot(xlim=(-x_axis-x_offset, x_axis-x_offset), ylim=(-y_axis-y_offset, y_axis-y_offset))
controlCell = mainGrid[1, :]
controlGrid = gridspec.GridSpecFromSubplotSpec(1, 7, controlCell)
sliderCell = controlGrid[0, 0]
sliderCount = 7
sliderGrid = gridspec.GridSpecFromSubplotSpec(sliderCount, 1, sliderCell)
sliders = []
for i in range(0, sliderCount):
    pass
    #sliders[i] = Slider(sliderGrid[0, i], "Test {}".format(i), 0.1, 8.0, valinit=2, valstep=0.01)

x_data = np.linspace(-x_axis-x_offset, x_axis-x_offset, 512)
y_data = [x for x in x_data]

line = plt.plot([], [])[0]
line.set_data(x_data, y_data)

plt.show()

Answer 1

Some issues:

• plot doesn't have any xlim argument.
• There is one grid too much in the code
• Widgets need to live inside axes
• The first index of a grid is the rows, not the columns.

In total,

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
from matplotlib.widgets import Slider

# Half width of the graph x-axis
x_axis = 4*np.pi
# x_axis offset
x_offset = 0
# Half height of the graph y-axis
y_axis = 8
# y_axis offset
y_offset = -1

fig = plt.figure()

mainGrid = gridspec.GridSpec(2, 1)
ax = plt.subplot(mainGrid[0, :])
ax.set(xlim=(-x_axis-x_offset, x_axis-x_offset), ylim=(-y_axis-y_offset, y_axis-y_offset))
controlCell = mainGrid[1, :]

sliderCount = 7
sliderGrid = gridspec.GridSpecFromSubplotSpec(sliderCount, 1, controlCell)
sliders = []
for i in range(0, sliderCount):
    sliderax = plt.subplot(sliderGrid[i, 0])
    slider = Slider(sliderax, "Test {}".format(i), 0.1, 8.0, valinit=2, valstep=0.01)
    sliders.append(slider)

x_data = np.linspace(-x_axis-x_offset, x_axis-x_offset, 512)
y_data = [x for x in x_data]

line = ax.plot([], [])[0]
line.set_data(x_data, y_data)

plt.show()