如何使用 pandas.to_excel() 创建 Excel **Table**？

Question

需要通过数据框以编程方式实现此目的：

https://docs.microsoft.com/en-us/power-bi/service-admin-troubleshoot-excel-workbook-data

Answer 1

这是使用 XlsxWriter 的一种方法：

import pandas as pd

# Create a Pandas dataframe from some data.
data = [10, 20, 30, 40, 50, 60, 70, 80]
df = pd.DataFrame({'Rank': data,
                   'Country': data,
                   'Population': data,
                   'Data1': data,
                   'Data2': data})

# Create a Pandas Excel writer using XlsxWriter as the engine.
writer = pd.ExcelWriter("pandas_table.xlsx", engine='xlsxwriter')

# Convert the dataframe to an XlsxWriter Excel object. Turn off the default
# header and index and skip one row to allow us to insert a user defined
# header.
df.to_excel(writer, sheet_name='Sheet1', startrow=1, header=False, index=False)

# Get the xlsxwriter workbook and worksheet objects.
workbook = writer.book
worksheet = writer.sheets['Sheet1']

# Get the dimensions of the dataframe.
(max_row, max_col) = df.shape

# Create a list of column headers, to use in add_table().
column_settings = []
for header in df.columns:
    column_settings.append({'header': header})

# Add the table.
worksheet.add_table(0, 0, max_row, max_col - 1, {'columns': column_settings})

# Make the columns wider for clarity.
worksheet.set_column(0, max_col - 1, 12)

# Close the Pandas Excel writer and output the Excel file.
writer.save()

输出：

更新：我在 XlsxWriter 文档中添加了一个类似的示例：Example: Pandas Excel output with a worksheet table

Answer 2

to_excel 无法做到这一点。一种解决方法是打开生成的 xlsx 文件并在其中添加带有openpyxl 的表：

import pandas as pd

df = pd.DataFrame({'Col1': [1,2,3], 'Col2': list('abc')})

filename = 'so58326392.xlsx'
sheetname = 'mySheet'
with pd.ExcelWriter(filename) as writer:
    if not df.index.name:
        df.index.name = 'Index'
    df.to_excel(writer, sheet_name=sheetname)

import openpyxl
wb = openpyxl.load_workbook(filename = filename)
tab = openpyxl.worksheet.table.Table(displayName="df", ref=f'A1:{chr(len(df.columns)+64)}{len(df)+1}')
wb[sheetname].add_table(tab)
wb.save(filename)

请注意所有表格标题必须是字符串。如果您有一个未命名的索引（这是规则），第一个单元格 (A1) 将为空，这会导致文件损坏。为避免这种情况，请为您的索引命名（如上所示）或使用以下方法导出不带索引的数据框：

df.to_excel(writer, sheet_name=sheetname, index=False)

Answer 3

如果您不想保存、重新打开和重新保存，另一种解决方法是使用xlsxwriter。它可以直接编写 ListObject 表，但不能直接从数据帧中编写，因此您需要拆分部分：

import pandas as pd
import xlsxwriter as xl

df = pd.DataFrame({'Col1': [1,2,3], 'Col2': list('abc')})

filename = 'output.xlsx'
sheetname = 'Table'
tablename = 'TEST'

(rows, cols) = df.shape
data = df.to_dict('split')['data']
headers = []
for col in df.columns:
    headers.append({'header':col})

wb = xl.Workbook(filename)
ws = wb.add_worksheet()

ws.add_table(0, 0, rows, cols-1,
    {'name': tablename
    ,'data': data
    ,'columns': headers})

wb.close()

add_table() 函数需要 'data' 作为列表列表，其中每个子列表代表数据帧的一行，'columns' 作为标题的字典列表，其中每一列由形成{'header': 'ColumnName'}。

Answer 4

我创建了一个包来从 pandas 编写格式正确的 excel 表格：pandas-xlsx-tables

from pandas_xlsx_tables import df_to_xlsx_table
import pandas as pd

data = [10, 20, 30, 40, 50, 60, 70, 80]
df = pd.DataFrame({'Rank': data,
                'Country': data,
                'Population': data,
                'Strings': [f"n{n}" for n in data],
                'Datetimes': [pd.Timestamp.now() for _ in range(len(data))]})

df_to_xlsx_table(df, "my_table", index=False, header_orientation="diagonal")

你也可以用xlsx_table_to_df做相反的事情