Images share both x- and y-axes

Question content

I have an image displayed using imshow. Then I add all rows and display the maximum value. I do the same thing with the columns. In the display graph, I want the x-axis and y-axis of the image to coincide with the x-axis of the added columns and the y-axis of the added rows. However, despite setting sharex and sharey respectively, it doesn't seem to work. I wish I could only do one at a time:

import numpy as np
import matplotlib.pyplot as plt
from scipy.signal import argrelextrema
import matplotlib.animation as animation

fig=        plt.figure()
gs=         fig.add_gridspec(2,2, height_ratios=[1, 0.1], width_ratios=[1, 0.1], hspace=0, wspace=0)
ax1=        fig.add_subplot(gs[0,0])
ax2=        fig.add_subplot(gs[1,0], sharex=ax1)
ax3=        fig.add_subplot(gs[0,1], sharey=ax1)
frameNumber= 10
imgs=   []

for i in range(frameNumber):
    np.random.seed(i)
    randomImage= np.random.random((5,5))
    sumX=   np.sum(randomImage, axis=0)
    sumY=   np.sum(randomImage, axis=1)
    dataRange=  np.arange(len(sumX))
    randomDataSet=  np.random.random((10))
    randomMaximalX= argrelextrema(sumX, np.greater)
    randomMaximalY= argrelextrema(sumY, np.greater)

    img1=   ax1.imshow(randomImage, animated=True)
    img2=   ax2.plot(dataRange, sumX,animated=True)[0]
    img3=   ax3.plot(sumY,dataRange,animated=True)[0]
    img4=   ax2.vlines(x=randomMaximalX, ymin=0, ymax=5, animated=True, linestyles="dashed")
    img5=   ax3.hlines(y=randomMaximalY, xmin=0, xmax=5, animated=True, linestyles="dashed")
    imgs.append([img1, img2, img3, img4, img5])
ani=    animation.ArtistAnimation(fig, imgs, interval=1000, blit=False)
plt.show()

The current result is this:

Actually I want something like this:

The h values ​​of the two graphs are the same. Thank you so much!

Correct answer

There are two ways to solve this problem:

1. Use axes.pcolormesh instead of axes.imshow
2. Or update the aspect ratio of adjacent images.

① axis.pcolormesh

axes.pcolormesh does not force the resulting image to be square (1:1 aspect ratio), so your cells will be rectangular, but they will fill the space provided appropriately.

from numpy.random import default_rng
import matplotlib.pyplot as plt

rng = default_rng(0)
image = rng.uniform(1, 10, size=(5, 5))

mosaic = [
    ['main',   'right'],
    ['bottom', '.'    ],
]

fig, axd = plt.subplot_mosaic(
    mosaic,
    gridspec_kw={
        'height_ratios': [1, .1], 'width_ratios': [1, .1],
        'wspace': .05, 'hspace': .05,
    },
    sharex=true,
    sharey=true,
)

axd['main'].pcolormesh(image)

plt.show()

② Update

If you want to stick with axes.imshow then you need to adjust Manually adjust the aspect ratio of each plot. To get the correct ratio you need Calculated based on height_ratio and width_ratio provided to gridspec

from numpy.random import default_rng
import matplotlib.pyplot as plt

rng = default_rng(0)
image = rng.uniform(1, 10, size=(5, 5))

mosaic = [
    ['main',   'right'],
    ['bottom', '.'    ],
]

fig, axd = plt.subplot_mosaic(
    mosaic,
    sharex=True,
    sharey=True,
    gridspec_kw={
        'height_ratios': [1, .1], 'width_ratios': [1, .1],

        # change values to move adjacent plots closer to the main
        'wspace': .05, 'hspace': .05,
    },
)

axd['main'].imshow(image)
axd['main'].set_anchor('SE') # move main plot to bottom-right of bounding-box

# calculate the width and height scales
gs = axd['main'].get_gridspec() # you can also save these values from your `gridspec_kw`
width_scale = gs.get_width_ratios()[0]    / gs.get_width_ratios()[1]
height_scale = gs.get_height_ratios()[0]  / gs.get_height_ratios()[1]

# update the aspect ratios of the adjacent plots
#   set their anchors so they correctly align with the main plot
axd['right'].set_aspect(width_scale, anchor='SW')
axd['bottom'].set_aspect(1/height_scale, anchor='NE')

plt.show()

The above is the detailed content of Images share both x- and y-axes. For more information, please follow other related articles on the PHP Chinese website!