Python Excel高级处理之数据透视表、条件格式、图表

pip install openpyxl  # Excel 高级功能（样式、图表、格式）

import pandas as pd

# 原始数据
df = pd.read_excel("销售数据.xlsx")
print(df.head())
# 输出：
#     日期    城市   销售员  品类    销售额
# 0  2026-01  北京   张三   手机    12000
# 1  2026-01  上海   李四   手机    15000
# 2  2026-01  北京   王五   电脑    18000

# 数据透视表：行=城市，列=品类，值=销售额总和
pivot = pd.pivot_table(
    df,
    values="销售额",
    index="城市",
    columns="品类",
    aggfunc="sum",
    fill_value=0,
    margins=True,        # 显示合计
    margins_name="合计"   # 合计列名
)

print(pivot)
# 品类     电脑     手机     合计
# 城市
# 北京   45000   32000   77000
# 上海   38000   41000   79000
# 合计   83000   73000  156000

pivot = pd.pivot_table(
    df,
    values="销售额",
    index=["城市", "销售员"],
    columns="品类",
    aggfunc="sum",
    fill_value=0,
)
# 行索引有两级：城市 → 销售员

with pd.ExcelWriter("销售报表.xlsx", engine="openpyxl") as writer:
    # 原始数据
    df.to_excel(writer, sheet_name="原始数据", index=False)
    # 透视表
    pivot.to_excel(writer, sheet_name="品类分析")
    # 多级透视表
    pivot_multi = pd.pivot_table(df, values="销售额",
        index=["城市", "销售员"], aggfunc="sum")
    pivot_multi.to_excel(writer, sheet_name="员工业绩")

from openpyxl import load_workbook
from openpyxl.styles import Font, PatternFill, Alignment, Border, Side

wb = load_workbook("销售报表.xlsx")

for sheet_name in wb.sheetnames:
    ws = wb[sheet_name]

    # 表头样式：加粗、居中、蓝底白字
    header_font = Font(bold=True, color="FFFFFF", size=11)
    header_fill = PatternFill(start_color="4472C4", end_color="4472C4", fill_type="solid")
    header_align = Alignment(horizontal="center", vertical="center")

    for cell in ws[1]:  # 第一行是表头
        cell.font = header_font
        cell.fill = header_fill
        cell.alignment = header_align

    # 数据行：居中，自动列宽
    for row in ws.iter_rows(min_row=2, max_col=ws.max_column):
        for cell in row:
            cell.alignment = Alignment(horizontal="center", vertical="center")

    # 自动列宽
    for col in ws.columns:
        max_length = 0
        col_letter = col[0].column_letter
        for cell in col:
            if cell.value:
                max_length = max(max_length, len(str(cell.value)))
        ws.column_dimensions[col_letter].width = max_length + 4

wb.save("销售报表_美化.xlsx")
print("样式美化完成")

from openpyxl.formatting.rule import CellIsRule, DataBarRule, ColorScaleRule
from openpyxl.styles import PatternFill

wb = load_workbook("销售报表.xlsx")
ws = wb.active

# 方法1：销售额 > 15000 标红
red_fill = PatternFill(start_color="FFC7CE", end_color="FFC7CE", fill_type="solid")
ws.conditional_formatting.add(
    "D2:D100",
    CellIsRule(operator="greaterThan", formula=["15000"], fill=red_fill))

# 方法2：数据条（直观显示大小）
ws.conditional_formatting.add(
    "D2:D100",
    DataBarRule(start_type="min", end_type="max",
                color="5B9BD5", showValue=True))

# 方法3：色阶（绿→黄→红）
ws.conditional_formatting.add(
    "D2:D100",
    ColorScaleRule(
        start_type="min", start_color="63BE7B",
        mid_type="percentile", mid_value=50, mid_color="FFEB84",
        end_type="max", end_color="F8696B"
    ))

wb.save("销售报表_条件格式.xlsx")

from openpyxl.chart import BarChart, Reference

wb = load_workbook("销售报表.xlsx")
ws = wb.active

# 创建柱状图
chart = BarChart()
chart.type = "col"          # 柱状图
chart.title = "各城市销售额"
chart.y_axis.title = "销售额"
chart.x_axis.title = "城市"
chart.style = 10

# 数据范围
data = Reference(ws, min_col=4, min_row=1, max_col=4, max_row=ws.max_row)
cats = Reference(ws, min_col=2, min_row=2, max_row=ws.max_row)  # 城市列

chart.add_data(data, titles_from_data=True)
chart.set_categories(cats)
chart.shape = 4

ws.add_chart(chart, "F2")
wb.save("销售报表_图表.xlsx")

from openpyxl.chart import LineChart

line_chart = LineChart()
line_chart.title = "月度销售趋势"
line_chart.style = 10
line_chart.y_axis.title = "销售额"
line_chart.x_axis.title = "月份"

data = Reference(ws, min_col=4, min_row=1, max_row=ws.max_row)
cats = Reference(ws, min_col=1, min_row=2, max_row=ws.max_row)

line_chart.add_data(data, titles_from_data=True)
line_chart.set_categories(cats)

ws.add_chart(line_chart, "F20")

from openpyxl.chart import PieChart

pie_chart = PieChart()
pie_chart.title = "品类占比"
pie_chart.style = 10

data = Reference(ws, min_col=4, min_row=1, max_row=ws.max_row)
cats = Reference(ws, min_col=3, min_row=2, max_row=ws.max_row)

pie_chart.add_data(data, titles_from_data=True)
pie_chart.set_categories(cats)

ws.add_chart(pie_chart, "F38")

import os
import pandas as pd

def merge_excel_files(input_dir, output_file):
    """合并整个文件夹的 Excel 文件"""
    all_data = []

    for f in os.listdir(input_dir):
        if f.endswith((".xlsx", ".xls")):
            df = pd.read_excel(os.path.join(input_dir, f))
            df["来源文件"] = f  # 标记来源
            all_data.append(df)
            print(f"已读取: {f} ({len(df)}行)")

    result = pd.concat(all_data, ignore_index=True)

    with pd.ExcelWriter(output_file, engine="openpyxl") as writer:
        result.to_excel(writer, sheet_name="汇总数据", index=False)

        # 生成报表
        pivot = pd.pivot_table(result,
            values="销售额", index="城市", aggfunc=["sum", "mean", "count"])
        pivot.to_excel(writer, sheet_name="统计报表")

    print(f"合并完成！共 {len(result)} 行 → {output_file}")

# 使用
merge_excel_files("月度报表", "年度汇总.xlsx")

import pandas as pd
from openpyxl import load_workbook
from openpyxl.chart import BarChart, Reference
from openpyxl.styles import Font, PatternFill, Alignment
from datetime import datetime

def generate_monthly_report(data_file, output_file):
    """自动生成月度销售报表"""

    # 1. 读取数据
    df = pd.read_excel(data_file)
    month = datetime.now().strftime("%Y年%m月")

    # 2. 生成透视表
    with pd.ExcelWriter(output_file, engine="openpyxl") as writer:
        # 原始数据
        df.to_excel(writer, sheet_name="数据源", index=False)

        # 按城市汇总
        city_pivot = pd.pivot_table(df,
            values="销售额", index="城市", aggfunc=["sum", "count"])
        city_pivot.to_excel(writer, sheet_name="城市分析")

        # 按品类汇总
        cat_pivot = pd.pivot_table(df,
            values="销售额", index="品类", aggfunc="sum")
        cat_pivot.to_excel(writer, sheet_name="品类分析")

    # 3. 美化 + 图表
    wb = load_workbook(output_file)
    ws = wb["城市分析"]

    # 表头样式
    header_font = Font(bold=True, color="FFFFFF", size=11)
    header_fill = PatternFill("solid", fgColor="4472C4")
    for cell in ws[1]:
        cell.font = header_font
        cell.fill = header_fill
        cell.alignment = Alignment(horizontal="center")

    # 柱状图
    chart = BarChart()
    chart.title = f"{month}各城市销售额"
    data = Reference(ws, min_col=2, min_row=1, max_col=2, max_row=ws.max_row)
    cats = Reference(ws, min_col=1, min_row=2, max_row=ws.max_row)
    chart.add_data(data, titles_from_data=True)
    chart.set_categories(cats)
    chart.width = 18; chart.height = 12
    ws.add_chart(chart, "D2")

    wb.save(output_file)
    print(f"月报已生成: {output_file}")

# 使用
generate_monthly_report("6月销售数据.xlsx", "6月销售报表.xlsx")

# 字体
Font(name="微软雅黑", size=11, bold=True, italic=False, color="FF0000")

# 填充
PatternFill(start_color="FFC000", end_color="FFC000", fill_type="solid")
GradientFill(stop=("FFFFFF", "4472C4"))

# 对齐
Alignment(horizontal="center", vertical="center", wrap_text=True)

# 边框
thin_border = Border(
    left=Side(style="thin"),
    right=Side(style="thin"),
    top=Side(style="thin"),
    bottom=Side(style="thin"),
)

# 行高列宽
ws.row_dimensions[1].height = 30
ws.column_dimensions["A"].width = 15