如何为多份材料测试CSV文件绘制应力应变平均线？

需求背景 #

我正在处理9份同类材料样本测试的CSV文件，已通过代码将所有样本的应力应变曲线绘制在同一张图中，现在希望添加一条平均线来对比同类材料的测试结果，减少图中曲线数量。不确定是生成存储平均值的新CSV文件，还是直接在现有循环中计算平均值，寻求最优方案。理想效果为在多条测试曲线之上叠加一条加粗的黑色平均线，作为所有样本的性能参考基准。

数据格式示例 #

Time,Displacement,Force,Flexure stress,Flexure strain (Displacement)
(s),(mm),(N),(MPa),(%)
"0.0000","0.0000","0.0007","0.0000","0.0000"
"0.0200","0.0000","0.0069","0.0004","0.0000"
"0.0400","0.0001","-0.0024","-0.0001","0.0003"
"0.0600","0.0005","0.0040","0.0002","0.0014"
"0.0800","0.0014","0.0106","0.0006","0.0041"

现有绘图代码 #

import os
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

### Set path to the folder containing the .csv files
PATH = 'my path' 

### Fetch all files in path
fileNames = os.listdir(PATH)

### Filter file name list for files ending with .csv
fileNames = [file for file in fileNames if '.csv' in file]

### Loop over all files
for file in fileNames:

    ### Read .csv file and append to list
    df = pd.read_csv(PATH + file, usecols = [3, 4], skiprows=2, names=['Stress', 'Strain'], header=None)
    
    strain_df = df['Strain']*0.01 
    side = 6.2 #mm
    stress_df = (df['Stress'])/(side**2) # N/mm**2

    ### Create line for every file
    plt.plot(strain_df, stress_df)

### Generate the plot
plt.xlabel(r'Strain $\epsilon$ (mm/mm)')
plt.ylabel(r'Stress $\sigma$ (N/mm$^2$)')
plt.title(r'Stress Strain Curve - 4$^\circ$C/min ')
plt.show()

方案建议 #

两种方式各有适用场景：

• 直接计算平均值绘图：如果仅需完成当前绘图需求，无需后续复用平均值数据，这种方式更高效，无需额外存储文件。
• 生成平均值CSV文件：如果需要后续对平均值数据进行分析、复现或其他用途，建议将平均值保存为CSV，方便后续调用。

方式1：直接计算并绘制平均线 #

修改现有代码，在循环中收集所有样本的应力数据，计算平均值后叠加绘图：

import os
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

PATH = 'my path' 
fileNames = [file for file in os.listdir(PATH) if '.csv' in file]

# 初始化列表存储所有样本的应力数据
all_stress = []
# 存储基准应变序列（取第一个文件的应变，假设所有文件应变点一致）
base_strain = None

for file in fileNames:
    df = pd.read_csv(PATH + file, usecols=[3,4], skiprows=2, names=['Stress','Strain'], header=None)
    strain_df = df['Strain']*0.01
    side = 6.2
    stress_df = df['Stress']/(side**2)
    
    plt.plot(strain_df, stress_df, alpha=0.5)  # 原曲线设为半透明，突出平均线
    
    # 存储应变和应力数据
    if base_strain is None:
        base_strain = strain_df.values
    all_stress.append(stress_df.values)

# 计算平均应力
mean_stress = np.mean(all_stress, axis=0)

# 绘制平均线
plt.plot(base_strain, mean_stress, color='black', linewidth=2, label='Average')

plt.xlabel(r'Strain $\epsilon$ (mm/mm)')
plt.ylabel(r'Stress $\sigma$ (N/mm$^2$)')
plt.title(r'Stress Strain Curve - 4$^\circ$C/min ')
plt.legend()
plt.show()

方式2：计算平均值并保存为CSV文件 #

在方式1的基础上，添加保存CSV的逻辑，后续可直接读取该文件绘图：

import os
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

PATH = 'my path' 
fileNames = [file for file in os.listdir(PATH) if '.csv' in file]

all_stress = []
base_strain = None

for file in fileNames:
    df = pd.read_csv(PATH + file, usecols=[3,4], skiprows=2, names=['Stress','Strain'], header=None)
    strain_df = df['Strain']*0.01
    side = 6.2
    stress_df = df['Stress']/(side**2)
    
    plt.plot(strain_df, stress_df, alpha=0.5)
    
    if base_strain is None:
        base_strain = strain_df.values
    all_stress.append(stress_df.values)

mean_stress = np.mean(all_stress, axis=0)

# 保存平均值到CSV
mean_df = pd.DataFrame({'Strain': base_strain, 'Stress': mean_stress})
mean_df.to_csv(os.path.join(PATH, 'average_stress_strain.csv'), index=False)

# 绘制平均线
plt.plot(base_strain, mean_stress, color='black', linewidth=2, label='Average')

plt.xlabel(r'Strain $\epsilon$ (mm/mm)')
plt.ylabel(r'Stress $\sigma$ (N/mm$^2$)')
plt.title(r'Stress Strain Curve - 4$^\circ$C/min ')
plt.legend()
plt.show()

注：如果不同文件的应变点数量或数值不一致，需先对所有样本的应力数据进行插值对齐（例如以统一的应变序列为基准，用np.interp插值得到对应应力），再计算平均值，确保平均线的准确性。

内容的提问来源于stack exchange，提问作者Anders Bleak