Python动态散点图无法显示散点仅更新刻度的问题排查

可正常运行的参考代码 #

这段代码能在同一张画布上实时更新散点图：

import matplotlib.pyplot as plt
import numpy as np
import time

# Create initial data
x = [15.1, 15.1, 15.1, 15.1, 15.1, 15.1, 15.1]
y =[10.3, 10.8, 11.3, 11.8, 12.3, 12.8, 13.3]
values = [31.628036009300377, 32.7041794935823, 32.41961219746959, 32.2074371367232, 32.69759838127627, 33.06824662635184, 32.44804260263164]  # Array of values for the colormap

plt.ion()  # Turn on interactive mode
fig, ax = plt.subplots()
scatter = ax.scatter(x, y, c=values, cmap='viridis', vmin=0, vmax=1)
ax.set_title("Dynamic Scatter Plot")

# Simulate continuous data updates in a loop
while True:
    # Generate new random data and values
    x = np.random.rand(50)
    y = np.random.rand(50)
    values = np.random.rand(50)

    # Update the scatter plot data and colors
    scatter.set_offsets(np.column_stack((x, y)))
    scatter.set_array(values)

    # Redraw the updated plot
    fig.canvas.draw()

    # Pause for a short duration to make the update visible
    plt.pause(0.1)

# Optionally, turn off interactive mode when you're done
plt.ioff()
plt.show()

我的问题代码 #

这段代码无法在初始画布上显示散点；取消注释循环内的4行代码后，每次循环会新建画布，而非更新原有画布：

import time, os,my_module
from my_module import getData,init_list_of_objects, plot
from statistics  import mean, stdev
import matplotlib.pyplot as plt
import numpy as np
from matplotlib.animation import FuncAnimation

global DATA
j=0
size=400 #(autodefine it using list dir somehow)

filename,x,y,z,labels,date=[],[],[],[],[],[]
average, deviation=[],[]

folder_path = "data"  # Relative folder path
  
def Statistics(data): #given an array with data, it does the maths for average and deviation
    average.append(mean(data))
    deviation.append(stdev(data))
    #print('\nAverage and deviation calculated.')

signal= init_list_of_objects(size) #defines a matrix for the signal 'cuz each file has an array of data--> I need a matrix for all the files
DATA=dict(filename=filename,x=x,y=y,signal=signal,labels=labels, date=date) #dictionary with all the data and info related to the excel files

seen_files = set()

plt.ion()  # Turn on interactive mode
fig, ax = plt.subplots()
scatter = ax.scatter(x, y, c=z, cmap='cool')
ax.set_title("Dynamic Scatter Plot")
# Create a color bar
cbar = fig.colorbar(scatter, ax=ax)

while True:
    new_files = [f for f in os.listdir(folder_path) if f.endswith('.xls') and f not in seen_files]
    for new_file in new_files:
        print(f"\n------------------------\nNew file added: {new_file},\n\t getting data...")
        getData(folder_path,new_file,DATA,j) 
        Statistics(DATA['signal'][j])
        
        x,y,z=plot(DATA,average)
        
        # Update the scatter plot data and colors
        scatter.set_offsets(np.column_stack((x, y)))
        scatter.set_array(z)

        # Update the color bar limits
        scatter.set_clim(vmin=min(z), vmax=max(z))

        # Redraw the updated plot
        fig.canvas.draw()
        fig.canvas.flush_events() 

        # Pause for a short duration to make the update visible
        plt.pause(1)
        
        '''fig, ax = plt.subplots()
        scatter = ax.scatter(x, y, c=z, cmap='cool')
        ax.set_title("Dynamic Scatter Plot")
        cbar = fig.colorbar(scatter, ax=ax)'''

           
        seen_files.add(new_file)
        j+=1
    time.sleep(10)  # Check every 10 seconds (adjust as needed)

plt.show()

核心疑问 #

为什么复用参考代码的逻辑，我的代码却无法更新初始画布？x、y、z数组会在循环中被正确填充，取消注释循环内的代码能生成正确图表，但每次都是新画布。尝试调整plt.show()或plt.clf()的位置也没用。

问题原因与修复方案 #

1. 初始数据为空导致的初始化失效 #

代码创建scatter对象时，x、y、z都是空列表：

filename,x,y,z,labels,date=[],[],[],[],[],[]
# ...
scatter = ax.scatter(x, y, c=z, cmap='cool')

空数据会让Matplotlib生成无效的散点对象，后续调用set_offsets和set_array无法正常触发显示更新。而参考代码一开始就有有效初始数据，所以能正常工作。

2. 颜色条未同步更新 #

即使更新了散点的颜色范围，颜色条不会自动刷新，需要手动同步。

修复步骤 #

第一步：初始化时用占位数据

替换空列表初始化，给x、y、z赋初始占位值：

# 替换原来的空列表初始化
x = [0]
y = [0]
z = [0]

plt.ion()
fig, ax = plt.subplots()
scatter = ax.scatter(x, y, c=z, cmap='cool')
ax.set_title("Dynamic Scatter Plot")
cbar = fig.colorbar(scatter, ax=ax)

第二步：更新循环中同步颜色条

在更新散点数据后，添加颜色条同步代码：

# Update the scatter plot data and colors
scatter.set_offsets(np.column_stack((x, y)))
scatter.set_array(z)

# Update the color bar limits and sync
scatter.set_clim(vmin=min(z), vmax=max(z))
cbar.update_normal(scatter)  # 新增这行同步颜色条

# Redraw the updated plot
fig.canvas.draw()
fig.canvas.flush_events() 

plt.pause(1)

额外注意事项 #

• 确保plot(DATA,average)返回的x、y、z是有效的可迭代对象，避免空数据。
• 全局变量j的递增不要超出signal矩阵的size=400限制，防止索引越界。

内容的提问来源于stack exchange，提问作者Dario