What This Tool Does

This tool extracts selected columns from your CSV file and creates a new CSV with only those columns. Choose columns by checking boxes next to their names, or specify column indices. Reorder columns by dragging or specifying the order. Optionally rename column headers in the output.

Selection Methods

By name: Check boxes next to column names to select them. Visual and intuitive for files with meaningful headers.

By index: Specify column positions (0-based or 1-based). Useful when you know exact positions or headers are missing.

Reordering: Drag columns to rearrange order, or specify the output order explicitly.

Renaming: Give columns new names in the output CSV without modifying the original file.

When to Use This

Data subsetting: Extract only the columns you need for analysis, reducing file size and complexity.

Privacy compliance: Remove sensitive columns (emails, phone numbers, IDs) before sharing datasets.

Feature selection: Prepare training data for machine learning by selecting only relevant features.

Report preparation: Create focused reports with only the columns stakeholders need to see.

Data transformation: Reorder columns to match a target system's expected column order.

Example

Input CSV (5 columns):

id,name,email,phone,department
1,Alice,[email protected],555-0100,Engineering
2,Bob,[email protected],555-0101,Marketing

Select: name, email, department

Output CSV (3 columns):

name,email,department
Alice,[email protected],Engineering
Bob,[email protected],Marketing

Column Index Reference

Column indices are 0-based (first column is 0, second is 1, etc.):

Column:     id    name    email    phone    department
Index:      0     1       2        3        4

To extract name, email, and department, you'd specify indices 1, 2, 4 (or 2, 3, 5 if using 1-based indexing).

Use Cases

GDPR compliance: A data analyst needs to share customer data with a contractor but must remove personal identifiers. They extract only purchase history columns, excluding names, emails, and addresses.

ML feature engineering: A data scientist has a 100-column dataset but only 15 features are relevant for their model. They extract those 15 columns to reduce training time.

API response mapping: An API expects data in a specific column order. A developer reorders columns from their database export to match the API specification.

Legacy system migration: An old system expects columns in a specific order with specific names. Columns are extracted, reordered, and renamed to match.

Limitations

Large files: Works best with files under 50MB. Very large files may cause slow performance in the browser.

Column count: Files with thousands of columns may be slow to render the selection interface.

Frequently Asked Questions

Can I extract non-contiguous columns?

Yes. Select any combination of columns regardless of their position in the original file.

What happens if I specify an index that doesn't exist?

Invalid indices are ignored. The tool extracts only columns that exist in the input file.

Can I duplicate columns in the output?

This tool doesn't support duplicating columns. Each selected column appears once in the output.