求助：将Matplotlib绘图插入Word文档表格时触发AttributeError错误

问题背景 #

开发模块需生成带3列表格的Word文档，调用子模块生成笛卡尔坐标系梯形图，但运行时触发错误：AttributeError: 'Figure' object has no attribute 'seek'

报错堆栈 #

File ~\anaconda3\Lib\site-packages\spyder_kernels\py3compat.py:356 in compat_exec
    exec(code, globals, locals)

File c:\users\barry\.spyder-py3\createtablewithspecificdimensions.py:40
    run.add_picture(fig, width=Inches(3.0), height=Inches(3.0))

File ~\anaconda3\Lib\site-packages\docx\text\run.py:79 in add_picture
    inline = self.part.new_pic_inline(image_path_or_stream, width, height)

File ~\anaconda3\Lib\site-packages\docx\parts\story.py:71 in new_pic_inline
    rId, image = self.get_or_add_image(image_descriptor)

File ~\anaconda3\Lib\site-packages\docx\parts\story.py:37 in get_or_add_image
    image_part = package.get_or_add_image_part(image_descriptor)

File ~\anaconda3\Lib\site-packages\docx\package.py:31 in get_or_add_image_part
    return self.image_parts.get_or_add_image_part(image_descriptor)

File ~\anaconda3\Lib\site-packages\docx\package.py:74 in get_or_add_image_part
    image = Image.from_file(image_descriptor)

File ~\anaconda3\Lib\site-packages\docx\image\image.py:49 in from_file
    stream.seek(0)

AttributeError: 'Figure' object has no attribute 'seek'

错误原因 #

run.add_picture()仅接受图片文件路径或二进制流对象，但当前传入的是matplotlib的Figure对象，该对象不具备文件流的seek()方法，导致报错。此外子模块已将图片保存为本地文件，直接使用文件路径即可。

修改方案 #

1. 被调用模块（Trapezoid_in_Cartesian_Grid_1.py）修改 #

调整函数返回值为图片文件路径，优化图形设置并释放内存：

from datetime import datetime
import matplotlib.pyplot as plt
import numpy as np
from matplotlib.path import Path
import matplotlib.patches as patches
from matplotlib.ticker import MultipleLocator
import matplotlib.ticker as ticker
import random

def ReflectTrap01OverXaxis(a,b,h,k):
    # 创建梯形顶点
    verts1 = [(a,b),(a+h,b),(a+h,b-k),(a,b-k+1),(a,b)]
    codes1 = [Path.MOVETO, Path.LINETO, Path.LINETO, Path.LINETO, Path.CLOSEPOLY]
    
    # 提前设置图尺寸，创建Figure
    fig = plt.figure(figsize=(3.0, 3.0))
    ax = fig.add_subplot(111)
    
    path1 = Path(verts1, codes1)
    patch1 = patches.PathPatch(path1, edgecolor="black", facecolor="None", lw=1, zorder=3)
    ax.add_patch(patch1)
    
    # 设置坐标轴与网格
    ax.set_xlim(-12,12)
    ax.set_ylim(-12,12)
    plt.grid(True)
    plt.tick_params(labelsize=6)
    plt.xlabel(None, fontsize=7)
    
    plt.axhline(color="black")
    plt.axvline(color="black")
    
    ax.grid(which="both", color="black", linewidth=0.4)
    ax.minorticks_on()
    
    # 生成唯一图片文件名并保存
    current_datetime = datetime.now().strftime("%Y-%m-%d %H%M%S")
    img_path = f"fig1_{current_datetime}.jpg"
    plt.savefig(img_path, dpi=1200, format="jpg")
    plt.close(fig)  # 关闭Figure释放内存
    
    return img_path  # 返回图片路径

2. 主模块修改 #

接收图片路径传入add_picture，修正表格单元格索引错误：

import os
import random
from docx import Document
from docx.shared import Inches
from datetime import datetime
import Trapezoid_in_Cartesian_Grid_1

doc = Document()

# 创建3列表格
tbl = doc.add_table(rows=1, cols=0)
tbl.add_column(width=Inches(0.4))
tbl.add_column(width=Inches(3.0))
tbl.add_column(width=Inches(3.0))

# 设置表头（可选）
header_row = tbl.rows[0]
header_row.cells[0].text = "序号"
header_row.cells[1].text = "题目要求"
header_row.cells[2].text = "图形"

for i in range(10):
    a = random.randint(-8,8)
    b = random.randint(-8,8)
    h = random.randint(-3,3)
    k = random.randint(-3,3)
    img_path = Trapezoid_in_Cartesian_Grid_1.ReflectTrap01OverXaxis(a,b,h,k)
    
    # 添加新行并填充内容
    row = tbl.add_row()
    row.cells[0].text = str(i+1) + "."
    row.cells[1].text = "Reflect the figure over the y-axis."
    
    # 向当前行第三列添加图片
    paragraph = row.cells[2].paragraphs[0]
    run = paragraph.add_run()
    run.add_picture(img_path, width=Inches(3.0), height=Inches(3.0))
    
    # 可选：删除临时图片，避免生成过多文件
    # os.remove(img_path)

# 生成文档文件名并设置表头
current_datetime = datetime.now().strftime("%Y-%m-%d %H%M%S")
file_name = f"WS_Proj_{current_datetime}"
header = doc.sections[0].header
header.paragraphs[0].text = file_name

doc.save(f"{file_name}.docx")

额外说明 #

• 被调用模块中，将plt.rcParams["figure.figsize"]改为创建Figure时直接指定figsize，确保尺寸设置生效。
• 添加plt.close(fig)避免多次循环后内存占用过高。
• 主模块修正了单元格引用错误，原代码会一直往表头单元格添加图片，现在改为当前行的对应单元格。
• 可启用os.remove(img_path)删除临时图片，减少文件冗余。

内容的提问来源于stack exchange，提问作者Mark Barry