Apply styles while exporting to 'xlsx' in pandas with XlsxWriter

Question

I use the .to_excel method of pandas to write a DataFrame as an Excel workbook. This works nice even for multi-index DataFrames as index cells become merged. When using the

Answer 1

The following approach allows me to use xlsxwriter formatting on the dataframe index and column names (though I can't guarantee it's validity):

import pandas as pd
import xlsxwriter as xl

# remove pandas header styles
# this avoids the restriction that xlsxwriter cannot
# format cells where formatting was already applied
pd.core.format.header_style = None

# write dataframe to worksheet
writer = pd.ExcelWriter(sumfile, engine='xlsxwriter')
df.to_excel(writer, sheet_name='test')

# create desired xlsxwriter formats
workbook  = writer.book
worksheet = writer.sheets['test']
header = workbook.add_format({'bold': True})
index = workbook.add_format({'align': 'left'})

# apply formats to header and index
worksheet.set_row(0, None, header)
worksheet.set_column(0,0, None, index)

Answer 2

Is there any way to do so

Currently no. There isn't a formatting mechanism like that in Pandas for formatting the Excel output (apart from a few hard-coded formats).

However, even if it was XlsxWriter doesn't currently support formatting cells after data is added. It is on TODO list.

Update:

As a workaround I recommend getting a reference to the underlying workbook and worksheet and overwriting any cells that you wish to be formatted with the same data from the Pandas dataframe and a XlsxWriter format.

See Working with Python Pandas and XlsxWriter.

Answer 3

If you just want to style the header, you can modify pandas.io.formats.excel.header_style. Of course, this is no general solution, but is an easy workaround for a common use-case.

import pandas.core.format
header_style_backup = pandas.io.formats.excel.header_style
try:
    pandas.io.formats.excel.header_style = {"font": {"bold": True},
                                       "borders": {"top": "thin", "right": "thin", "bottom": "thin", "left": "thin"},
                                       "pattern": {"pattern": "solid", "fore_colour": 26},
                                       "alignment": {"horizontal": "center", "vertical": "top"}}
    df.to_excel(writer, sheet_name=sheetname, startrow=table_startrow)
finally:
    pandas.formats.format.header_style = header_style_backup

Note: The location of header_style has been changing multiple times in prior pandas versions. Use the following for older versions:

version < 0.20.0 pandas.formats.format.header_style

version < 0.18.0 pandas.core.format.header_style

Answer 4

The next version of Pandas (2.0) will include experimental support for exporting styled DataFrames direct to Excel using openpyxl: http://pandas-docs.github.io/pandas-docs-travis/style.html#Export-to-Excel