Analyze data

Upload a spreadsheet to Le Chat ↗ and ask questions about it in plain language.

• Code Interpreter runs Python in a secure sandbox to produce charts, tables, and statistical summaries.
• You don't write any code; the model handles pandas, matplotlib, and formatting.

Time to complete: ~10 minutes

• A Le Chat account (Pro or Team plan recommended; Free tier has daily limits).
• A data file to analyze (CSV, XLSX, or JSON)

1. Open New chat ↗.
2. Click the attachment icon (paperclip) in the message bar.
3. Select your data file: for example, sales-q4-2025.csv.
4. Le Chat displays a preview of the file. Confirm it looks correct.

For best results, use files with clear column headers. The model reads them to understand the data structure.

Type a question in natural language. Code Interpreter runs Python (pandas, matplotlib) in a secure sandbox and returns the result.

Example prompts to try:

Summarize this dataset. How many rows and columns are there? What are the key statistics?

Show me monthly revenue trends as a line chart.

What are the top 5 products by total sales? Display as a bar chart.

Calculate the correlation between marketing spend and revenue.

The model writes and runs Python code automatically. You see tables, charts, and text directly in the chat.

Build on previous results by asking follow-up questions. The model remembers the data context.

1. Refine a chart: "Make the chart wider and add data labels."
2. Filter data: "Show only rows where region is Europe."
3. Compare periods: "Compare Q3 vs Q4 revenue by product category."

To save your results:

1. Download charts: right-click any generated chart and select Save image.
2. Copy tables: highlight and copy any generated table.
3. Use Canvas: ask "Put this summary in a Canvas" to create an editable document with your analysis.

Your data analysis is working correctly if:

• The model correctly identified your column names and data types
• Generated charts display accurate data from your file
• Follow-up questions reference the same dataset without re-uploading
• Statistical calculations (mean, median, correlation) match expected values

If the model misinterprets a column, try: "The 'Date' column is in DD/MM/YYYY format" to clarify.