将多个 matplotlib 图形合并为一个

Question

I have a function that takes an enhanced dicom file and does the following:

1. Use a for loop that creates a single slice from the dicom file and then index it into a smaller array around a set of six specks.
2. A second for loop that draws circles around the six specks and one for the background
3. Show the image with all the circles plotted and the current slice location.

When I call the function I can create multiple images that are displayed in separate figures. So, how can I combine all these figures into one? Ideally I want to display them in a 3x3 grid.

Here is my current code:

import matplotlib.pyplot as plt
import pydicom # Used for opening DICOM files
import numpy as np # General mathematical package
from pylab import text

# Import DICOM files
filename = "U:/File location" 
ds = pydicom.dcmread(filename)

# Speck Locations
centerSpeck = (1690, 1477)
centerSpeck2 = (100, 100)
twelveOclockSpeck = (45, 84)
twoOkclockSpeck = (66, 148)
tenOclockSpeck = (136, 147)
fiveOclockSpeck = (157, 82)
sevenOclockSpeck = (102, 41)
backgroundSignalValue = (71, 63)

speckLocations = np.array([centerSpeck2, twelveOclockSpeck, twoOkclockSpeck, fiveOclockSpeck, sevenOclockSpeck, tenOclockSpeck, backgroundSignalValue])

# Function that draws a circle around a given pixel
def drawCircle(arrayToPLot, zeroIndex, oneIndex, specks, r = 10):
    x = speckLocations[:,1] # get x axis variables from the speckLocations array
    y = speckLocations[:,0] # get y axis variables from the speckLocations array
    for i in range(zeroIndex,oneIndex,1): # For loop, note that the function range is: range(start, stop, step)
        tempIm = arrayToPLot.pixel_array[i,:,:].astype(float) # Get one slice as float
        slicedArray = tempIm[centerSpeck[0]-100:centerSpeck[0]+100, centerSpeck[1]-100:centerSpeck[1]+100].astype(float)
        for x,y in (speckLocations):
            plt.imshow(slicedArray, cmap='gray')
            circle = plt.Circle((y, x),r, fc='none', ec="red")
            plt.gca().add_patch(circle)    
            text(10, 180, i, fontsize=12, color='red') # Print the current slice on the image

        plt.show() # Plot each slice with circles drawn around all the specks and the background signal value location

#Call the function  
drawCircle(ds, 33, 34, speckLocations)

Answer 1

In the following we'll use the plt.subplots method to produce a figure and a grid of axes , the Matplotlib objects that someone understands as subplots ...

Iteration on a 2D grid first give you a row of the grid, a second iteration on the row gives you the individual subplot. When we have singled out an axes, it's time to call your function (you have the responsibility to differentiate the data to plot because it's not clear from your question what you want to do) but with an additional argument, the current axes.

fig, ax_grid = plt.subplots(3,3)
for ax_row in ax_grid:
    for ax in ax_row:
        drawCircle(ax, ds, 33, 34, speckLocations)
plt.tight_layout()
plt.show()

At the end we show the figure, calling tight_layout to get a better arrangement of the subplots.

---

Of course we have to modify the circle plotting function... first we add ax to the list of arguments, next we modify the calls to plt methods to use the methods of the ax object:

def drawCircle(ax, arrayToPLot, zeroIndex, oneIndex, specks, r = 10):
    ...
    ax.imshow(...)
    ...
    ax.add_patch(circle)
    ax.text(10, 180, i, fontsize=12, color='red')