基于Python实现帕累托图的示例详解

# 用于命名和保存图片文件
import os
from datetime import datetime


import matplotlib.font_manager as fm  # 管理字体工具
import matplotlib.pyplot as plt  # 绘图包
import pandas as pd # 读取Excel数据

# 设置字体
font_path = 'C:/Windows/Fonts/simhei.ttf'  # 字体文件路径
font_prop = fm.FontProperties(fname=font_path, size=12)  # 字体属性
plt.rcParams['font.family'] = font_prop.get_name()

# 读取Excel数据
df = pd.read_excel('data.xlsx')

# 建立数据category和标识符的对应关系
category_dict = {}
for i, category in enumerate(df['category'].unique()):
    category_dict[category] = chr(65 + i)
    df['category'] = df['category'].replace(category, chr(65 + i))

# 按照数量降序排列
df = df.sort_values(by='value', ascending=False)

# 计算累计百分比
df['cumulative_percentage'] = df['value'].cumsum() / df['value'].sum() * 100

# 绘制帕累托图
fig, ax1 = plt.subplots(figsize=(8, 10))

# 判断是否有小于80%的数据
if df['cumulative_percentage'].min() < 80:
    # 将小于80%的数据用红色柱子表示
    ax1.bar(df[df['cumulative_percentage'] < 80].index, df[df['cumulative_percentage'] < 80]['value'], color='tab:red')
else:
    # 将第一个柱子设为红色背景
    ax1.bar(df.index[0], df['value'][0], color='tab:red')

# 绘制其余柱子
bar_heights = ax1.bar(df.index, df['value'], color='tab:blue', alpha=0.5)
ax1.set_ylabel('数量', fontproperties=font_prop)

# 在每个柱子上方添加具体的值
for i, bar_height in enumerate(bar_heights):
    ax1.text(i, bar_height.get_height() + 0.5, str(int(bar_height.get_height())), ha='center', fontproperties=font_prop)

ax2 = ax1.twinx()
ax2.plot(df.index, df['cumulative_percentage'], color='tab:red', marker='o')
ax2.set_ylim([0, 100])
ax2.set_ylabel('累计百分比', fontproperties=font_prop)

plt.xticks(df.index, df['category'], rotation=90, fontproperties=font_prop)
plt.title('帕累托图', fontproperties=font_prop)

# 添加A-Z和category的对应关系
ax1.set_xticks(df.index)
ax1.set_xticklabels(df['category'], rotation=90, fontproperties=font_prop)
ax1.tick_params(axis='x', which='major', pad=15)
ax1.spines['bottom'].set_position(('axes', -0.00))
ax1.spines['bottom'].set_linewidth(0)
ax1.spines['bottom'].set_color('gray')
ax1.spines['bottom'].set_visible(True)

for i, (category, abbr) in enumerate(category_dict.items()):
    ax1.text(-1, -4 * i - 10, f'{abbr}: {category}', transform=ax1.transData, ha='left', fontproperties=font_prop,
             va='top')

# 调整子图之间的间距和边距
plt.subplots_adjust(bottom=0.4)

# 保存图片
now = datetime.now().strftime('%Y%m%d%H%M%S')
filename = os.path.join(os.path.dirname(os.path.abspath(__file__)), f'pareto_{now}.png')
plt.savefig(filename)

plt.show()