pandas xlsxwriter, format header

Question

I'm saving pandas DataFrame to_excel using xlsxwriter. I've managed to format all of my data (set column width, font size etc) except for changing header's font and I can't find the way to do it. Here's my example:

import pandas as pd
data = pd.DataFrame({'test_data': [1,2,3,4,5]})
writer = pd.ExcelWriter('test.xlsx', engine='xlsxwriter')

data.to_excel(writer, sheet_name='test', index=False)

workbook  = writer.book
worksheet = writer.sheets['test']

font_fmt = workbook.add_format({'font_name': 'Arial', 'font_size': 10})
header_fmt = workbook.add_format({'font_name': 'Arial', 'font_size': 10, 'bold': True})

worksheet.set_column('A:A', None, font_fmt)
worksheet.set_row(0, None, header_fmt)

writer.save()

The penultimate line that tries to set format for the header does nothing.

Answer 1

An update for anyone who comes across this post and is using Pandas 0.20.1.

It seems the required code is now

import pandas.io.formats.excel
pandas.io.formats.excel.header_style = None

Apparently the excel submodule isn't imported automatically, so simply trying pandas.io.formats.excel.header_style = None alone will raise an AttributeError.

Answer 2

I think you need first reset default header style, then you can change it:

pd.core.format.header_style = None

All together:

import pandas as pd

data = pd.DataFrame({'test_data': [1,2,3,4,5]})
writer = pd.ExcelWriter('test.xlsx', engine='xlsxwriter')

pd.core.format.header_style = None

data.to_excel(writer, sheet_name='test', index=False)

workbook  = writer.book
worksheet = writer.sheets['test']

font_fmt = workbook.add_format({'font_name': 'Arial', 'font_size': 10})
header_fmt = workbook.add_format({'font_name': 'Arial', 'font_size': 10, 'bold': True})

worksheet.set_column('A:A', None, font_fmt)
worksheet.set_row(0, None, header_fmt)

writer.save()

Explaining by jmcnamara, thank you:

In Excel a cell format overrides a row format overrides a column format.The pd.core.format.header_style is converted to a format and is applied to each cell in the header. As such the default cannot be overridden by set_row(). Setting pd.core.format.header_style to None means that the header cells don't have a user defined format and thus it can be overridden by set_row().

EDIT: In version 0.18.1 you have to change

pd.core.format.header_style = None

to:

pd.formats.format.header_style = None

thanks krvkir.

Answer 3

In pandas 0.20 the solution of the accepted answer changed again.

The format that should be set to None can be found at:

pandas.io.formats.excel.header_style