Exporting data and charts

to_csv, to_excel, multiple sheets, and saving figures properly.

What you'll be able to do

Export to CSV and Excel
Write several sheets to one workbook
Save charts at usable quality

CSV

df.to_csv('output.csv', index=False)

index=False

Without it, pandas writes the index as an unnamed first column. Re-import that file and you get a stray Unnamed: 0 column — and it compounds every round trip.

df.to_csv('output.csv', index=False, encoding='utf-8-sig')

utf-8-sig makes Excel display accented characters correctly rather than as mojibake.

Excel

df.to_excel('report.xlsx', index=False, sheet_name='Summary')

Several sheets in one workbook:

with pd.ExcelWriter('report.xlsx') as writer:
    summary.to_excel(writer, sheet_name='Summary', index=False)
    detail.to_excel(writer,  sheet_name='Detail',  index=False)
    by_region.to_excel(writer, sheet_name='Regions', index=False)

The with block handles saving and closing — a common way to end up with a corrupt file is writing sheets without it.

Figures

fig.savefig('chart.png', dpi=150, bbox_inches='tight')
fig.savefig('chart.svg', bbox_inches='tight')       # vector, scales cleanly

dpi=150 is a reasonable screen and document default; 300 for print. SVG stays sharp at any size, which is the better choice for anything going into a slide deck.

Rounding before export

df.round({'revenue': 2, 'margin': 3}).to_csv('output.csv', index=False)

Saves whoever opens it from 1250.7599999999998, which looks like carelessness even though it is just floating point.

Key points

Always pass index=False to to_csv unless the index is meaningful
Use ExcelWriter in a with block for multi-sheet workbooks
SVG for slides, PNG at dpi=150 for documents

Check yourself