使用 xlsxwriter 将 pandas 数据帧写入 Excel 并包含“write_rich_string”格式

Question

以下内容可重现并生成所需的输出。

import xlsxwriter, pandas as pd

workbook = xlsxwriter.Workbook('pandas_with_rich_strings.xlsx')
worksheet = workbook.add_worksheet()

# Set up some formats to use.
bold = workbook.add_format({'bold': True})
italic = workbook.add_format({'italic': True})
red = workbook.add_format({'color': 'red'})

df = pd.DataFrame({
    'numCol': [1, 50, 327],
    'plainText': ['plain', 'text', 'column'],
    'richText': [
        ['This is ', bold, 'bold'],
        ['This is ', italic, 'italic'],
        ['This is ', red, 'red']
    ]
}) 


headRows = 1


for colNum in range(len(df.columns)):
    xlColCont = df[df.columns[colNum]].tolist()
    worksheet.write_string(0, colNum , str(df.columns[colNum]), bold)
    for rowNum in range(len(xlColCont)):
        if df.columns[colNum] == 'numCol': 
            worksheet.write_number(rowNum+headRows, colNum , xlColCont[rowNum])        
        elif df.columns[colNum] == 'richText':
            worksheet.write_rich_string(rowNum+headRows, colNum , *xlColCont[rowNum])
        else:
            worksheet.write_string(rowNum+headRows, colNum , str(xlColCont[rowNum]))


workbook.close()

但是，如果不遍历每一列并一次性将整个 pandas 数据帧写入 Excel 文件，我将如何做同样的事情并且包括write_rich_string格式？

以下无效。

writer = pd.ExcelWriter('pandas_with_rich_strings.xlsx', engine='xlsxwriter')
workbook = xlsxwriter.Workbook('pandas_with_rich_strings.xlsx')
worksheet = workbook.add_worksheet('pandas_df')
df.to_excel(writer,'pandas_df')
writer.save()

Answer 1

我不确定我的回答是否比您的回答好得多，但我已将其缩减为仅使用一个 for 循环并利用 pandas.DataFrame.to_excel() 最初放置数据框在 excel。请注意，我随后使用 worksheet.write_rich_string().

覆盖了最后一列

import pandas as pd

writer = pd.ExcelWriter('pandas_with_rich_strings.xlsx', engine='xlsxwriter')
workbook  = writer.book
bold = workbook.add_format({'bold': True})
italic = workbook.add_format({'italic': True})
red = workbook.add_format({'color': 'red'})
df = pd.DataFrame({
    'numCol': [1, 50, 327],
    'plainText': ['plain', 'text', 'column'],
    'richText': [
        ['This is ', bold, 'bold'],
        ['This is ', italic, 'italic'],
        ['This is ', red, 'red']
    ]
}) 
df.to_excel(writer, sheet_name='Sheet1', index=False)
worksheet = writer.sheets['Sheet1']
# you then need to overwite the richtext column with
for idx, x in df['richText'].iteritems():
    worksheet.write_rich_string(idx + 1, 2, *x)
writer.save()

预期输出 .xlsx：

Answer 2

您的代码可以通过使用 worksheet.add_write_handler() 检测列表并从 worksheet.write 自动调用 worksheet.write_rich_string() 而无需手动检查类型来简化。你会想

worksheet.add_write_handler(list, xlsxwriter.worksheet.Worksheet.write_rich_string)

应该可以，但不能，因为该方法的可变参数有一些问题被破坏了（最后一个参数是整个单元格的可选样式）。因此，以下确实有效

worksheet.add_write_handler(list, lambda worksheet, row, col, args: worksheet._write_rich_string(row, col, *args))

不幸的是，这种方法不容易与 pd.to_excel 兼容，因为它必须在写入数据之前在工作表上设置，并且因为 ExcelWriter 在写入之前将列表和字典序列化为字符串（有一个注释在说明这是为了与 CSV 编写器兼容的文档中）。 Subclassing pd.io.excel._xlsxwriter._XlsxWriter 可以工作：

import xlsxwriter, pandas as pd

class RichExcelWriter(pd.io.excel._xlsxwriter._XlsxWriter):
    def __init__(self, *args, **kwargs):
        super(RichExcelWriter, self).__init__(*args, **kwargs)

    def _value_with_fmt(self, val):
        if type(val) == list:
            return val, None
        return super(RichExcelWriter, self)._value_with_fmt(val)

    def write_cells(self, cells, sheet_name=None, startrow=0, startcol=0, freeze_panes=None):
        sheet_name = self._get_sheet_name(sheet_name)
        if sheet_name in self.sheets:
            wks = self.sheets[sheet_name]
        else:
            wks = self.book.add_worksheet(sheet_name)
            #add handler to the worksheet when it's created
            wks.add_write_handler(list, lambda worksheet, row, col, list, style: worksheet._write_rich_string(row, col, *list))
            self.sheets[sheet_name] = wks
        super(RichExcelWriter, self).write_cells(cells, sheet_name, startrow, startcol, freeze_panes)



writer = RichExcelWriter('pandas_with_rich_strings_class.xlsx')
workbook  = writer.book
bold = workbook.add_format({'bold': True})
italic = workbook.add_format({'italic': True})
red = workbook.add_format({'color': 'red'})
df = pd.DataFrame({
    'numCol': [1, 50, 327],
    'plainText': ['plain', 'text', 'column'],
    'richText': [
        ['This is ', bold, 'bold'],
        ['This is ', italic, 'italic'],
        ['This is ', red, 'red']
    ]
})

df.to_excel(writer, sheet_name='Sheet1', index=False)

writer.save()

或者我们可以直接使用 xlsxwriter 并使用来自 pandas 的 ExcelFormatter，它也处理 header 格式并采用许多与 to_excel 相同的参数。

import xlsxwriter, pandas as pd
from pandas.io.formats.excel import ExcelFormatter

workbook = xlsxwriter.Workbook('pandas_with_rich_strings.xlsx')
worksheet = workbook.add_worksheet()

# Set up some formats to use.
bold = workbook.add_format({'bold': True})
italic = workbook.add_format({'italic': True})
red = workbook.add_format({'color': 'red'})

df = pd.DataFrame({
    'numCol': [1, 50, 327],
    'plainText': ['plain', 'text', 'column'],
    'richText': [
        ['This is ', bold, 'bold'],
        ['This is ', italic, 'italic'],
        ['This is ', red, 'red']
    ]
})

worksheet.add_write_handler(list, lambda worksheet, row, col, args: worksheet.write_rich_string(row, col, *args))
cells = ExcelFormatter(df, index=False).get_formatted_cells()

for cell in cells:
    worksheet.write(cell.row, cell.col ,cell.val)

workbook.close()

这会产生所需的输出，无需对数据进行两次循环。事实上，它使用与 pandas 相同的生成器，因此它与 pandas.to_excel() 一样高效，并且格式化程序 class 采用许多相同的参数。

使用 xlsxwriter 将 pandas 数据帧写入 Excel 并包含“write_rich_string”格式

Write pandas dataframe to Excel with xlsxwriter and include `write_rich_string` formatting

python

python-3.x

pandas

xlsxwriter