Analyzing Spreadsheets

When to use

• User shares an Excel workbook or asks about spreadsheet analysis
• Tasks include summarizing metrics, spotting anomalies, or drafting charts
• Data lives in tabular form (CSV or XLSX)

Workflow

1. Inspect workbook structure

   import pandas as pd
   xl = pd.ExcelFile("input.xlsx")
   xl.sheet_names
   

2. Load relevant sheets

   df = pd.read_excel("input.xlsx", sheet_name="Sheet1")
   df.head()
   

3. Clean and validate
   • Drop empty columns/rows
   • Normalize date formats with pd.to_datetime
   • Verify numeric columns with df.describe()
4. Analyze and summarize
   • Use groupby/pivot patterns from reference/pandas-recipes.md
   • Highlight KPIs, trends, and outliers
5. Recommend visuals
   • Suggest chart types (line for time series, bar for categorical comparisons, heatmap for correlations)
   • Provide short rationale per recommendation

Output expectations

• Concise summary (1–3 paragraphs) covering key findings
• Bullet list of insights with supporting numbers
• Optional chart suggestions with column mappings

Validation checklist

• Loaded the correct sheet(s) and reported row/column counts
• Highlighted missing or unusual data
• Referenced actual values from the workbook
• Included next-step recommendations (e.g., further slicing, charting)

Additional resources

• reference/pandas-recipes.md – common aggregation patterns
• python -m pip install pandas openpyxl – install requirements if missing (Claude Code already includes pandas)