python xlsxwriter: Keep header in excel when adding a table

Question

I have a panda dataframe that I write to a xslx file, and would like to add a table over that data. I would also like to keep the headers that I have already written, instead of adding them again. Is that possible?

Example:

import pandas as pd
import xlsxwriter as xw

# random dataframe
d = {'one' : pd.Series([1., 2., 3.], index=['a', 'b', 'c']), 'two' : pd.Series([5., 6., 7., 8.], index=['a', 'b', 'c', 'd'])}
df = pd.DataFrame(d)


# write data to file
writer = pd.ExcelWriter("test.xlsx", engine='xlsxwriter')
df.to_excel(writer,"sheet without table")
df.to_excel(writer,"sheet with table")
df.to_excel(writer,"sheet with table and header")

# get sheets to add the tables
workbook  = writer.book
worksheet_table = writer.sheets['sheet with table']
worksheet_table_header = writer.sheets['sheet with table and header']

# the range in which the table is
end_row = len(df.index)
end_column = len(df.columns)
cell_range = xw.utility.xl_range(0, 0, end_row, end_column)


# add the table that will delete the headers
worksheet_table.add_table(cell_range,{'header_row': True,'first_column': True})

######################################
# The hack

# Using the index in the Table
df.reset_index(inplace=True)
header = [{'header': di} for di in df.columns.tolist()]
worksheet_table_header.add_table(cell_range,{'header_row': True,'first_column': True,'columns':header})

writer.save()

Answer 1

The hack / work around is the only option (as seen from @jmcnamara). In short it is:

import pandas as pd
import xlsxwriter as xw

# random dataframe
d = {'one' : pd.Series([1., 2., 3.], index=['a', 'b', 'c']), 'two' : pd.Series([5., 6., 7., 8.], index=['a', 'b', 'c', 'd'])}
df = pd.DataFrame(d)


# write data to file
writer = pd.ExcelWriter("test.xlsx", engine='xlsxwriter')
df.to_excel(writer,"sheet with table and header")

# get sheets to add the tables
workbook  = writer.book
worksheet_table_header = writer.sheets['sheet with table and header']

# the range in which the table is
end_row = len(df.index)
end_column = len(df.columns)
cell_range = xw.utility.xl_range(0, 0, end_row, end_column)

######################################
# The hack

# Using the index in the Table
df.reset_index(inplace=True)
header = [{'header': di} for di in df.columns.tolist()]
worksheet_table_header.add_table(cell_range,{'header_row': True,'first_column': True,'columns':header})

writer.save()

Answer 2

How about this (note that the 'options' is only required if the data frame contains NAs):

import pandas as pd
import xlsxwriter

# random dataframe
d = {'one':pd.Series([1., 2., 3.]), 'two':pd.Series([5., 6., 7., 8.])}
df = pd.DataFrame(d)

workbook = xlsxwriter.Workbook('test.xlsx', options={'nan_inf_to_errors': True})
worksheet = workbook.add_worksheet('sheet1')
worksheet.add_table(0, 0, df.shape[0], df.shape[1]-1,
    {'data': df.values.tolist(),
    'columns': [{'header': c} for c in df.columns.tolist()],
    'style': 'Table Style Medium 9'})
workbook.close()

Answer 3

I would also like to keep the headers that I have already written, instead of adding them again. Is that possible?

No.

Your third solution in worksheet_table_header is probably the best way to implement it.

Answer 4

I had to modify the the hack by @jmcnamara when using xlsxwriter version 0.9.6. I had to subtract one from the number of columns or I ended up with an extra column that was not in the pandas.DataFrame (see end_column assignment). Modified version below (pandas version 0.19.2).

import pandas as pd
import xlsxwriter

# random dataframe
d = {'one':pd.Series([1., 2., 3.]), 'two':pd.Series([5., 6., 7., 8.])}
df = pd.DataFrame(d)
print df

# write data to file
writer = pd.ExcelWriter("test.xlsx", engine='xlsxwriter')
df.to_excel(writer, 'sheet1', index=False)

# get sheets to add the tables
ws = writer.sheets['sheet1']

# the range in which the table is
end_row = len(df.index)
end_column = len(df.columns) - 1
cell_range = xlsxwriter.utility.xl_range(0, 0, end_row, end_column)

######################################
# The hack
header = [{'header': c} for c in df.columns.tolist()]
ws.add_table(cell_range,{'header_row': True, 'columns':header, 'style':'Table Style Medium 11'})
ws.freeze_panes(1, 1)
writer.save()
writer.close()