Learning Timeline
Key Insights
Prompt 'Retrofitting' Tips
You should customize (retrofit) the niche section of the prompt based on your target industry. If you are looking for small businesses, ensure you instruct Claude to exclude big-box retailers from the dataset.
Advantages of Claude Code vs. Manual Work
Claude Code is highly efficient because it can process multiple CSV files in bulk without needing to open spreadsheet applications like Excel or perform manual filtering one by one.
Prompts
CSV Junk Data Cleaning Prompt
Target:
Claude Code
I have these CSV files in my directory. Go ahead and clean them by removing all of the obvious junk data. Remove things like listings with no business name, address, city, or state. Also, remove permanently closed ones, and any obvious ones that don't relate to my niche like big box retailers. Process all the files and provide the cleaned version.
Step by Step
Steps to Clean CSV Datasets with Claude Code
- Open your terminal or Command Line Interface (CLI) in the project directory containing your CSV files.
- Ensure Claude Code is installed and ready to accept commands within that directory.
- Identify the CSV files that need cleaning (e.g., 5 files from web scraping results).
- Enter a cleaning prompt into Claude Code, specifying the exact data criteria you want to remove.
- Instruct Claude to filter out rows with missing information in the 'business name', 'address', 'city', or 'state' columns.
- Add instructions to remove entities marked as 'permanently closed'.
- Specify irrelevant business categories (e.g., 'big box retailers') to be excluded from the list.
- Let Claude Code process all CSV files simultaneously to update the data.
- Review the summary or final file output to ensure only high-quality data remains before starting your analysis.
