Does shuffling columns change any of the data values?

No, every column moves as a unit: its header and every data cell in it travel together, so the values within a column are never altered, only the column's position.

How is the randomness generated?

With a Fisher-Yates shuffle driven by crypto.getRandomValues() rather than Math.random(), the same rejection-sampling approach the random category's other tools use for unbiased results.

How is this different from CSV Row Shuffler?

This tool reorders columns (left-to-right); CSV Row Shuffler reorders data rows (top-to-bottom) instead, leaving the header fixed at the top.

Can I get the same shuffle order twice?

No, each run generates a fresh cryptographically random order; the tool doesn't support seeding or reproducing a previous shuffle.

CSV Column Shuffler - Randomize Column Order

Overview

Introduction

Sometimes you need to randomize the order columns appear in, for testing that a downstream system doesn't assume a fixed column order, or for creating varied sample data. This tool shuffles whole columns at once.

It uses a cryptographically strong random source so the shuffle isn't biased toward any particular ordering.

What Is CSV Column Shuffler?

A whole-column shuffler for CSV: it randomly reorders which position each column appears in, moving the header cell and every data cell in that column together as one unit.

Randomness comes from a Fisher-Yates shuffle seeded by crypto.getRandomValues(), the same rejection-sampling approach used elsewhere in this repo's random category.

How CSV Column Shuffler Works

The CSV is parsed and padded to a rectangle. A list of column indices is generated and shuffled in place with Fisher-Yates, drawing each swap index from crypto.getRandomValues() with rejection sampling to avoid modulo bias.

Every row (including the header) is then rebuilt by reading its cells in the new shuffled column order.

When To Use CSV Column Shuffler

Use it to generate test fixtures that verify a system doesn't secretly depend on column order.

It's also handy for creating visually varied sample data from a single source CSV.

Often used alongside CSV Row Shuffler and CSV Value Shuffler.

Features

Advantages

Uses a cryptographically strong random source rather than Math.random for an unbiased shuffle.
Keeps each column's own values fully intact: only column position changes.
Works on the whole grid at once, header and data rows included, in a single pass.

Limitations

Requires at least two columns to produce a meaningful shuffle.
There's no way to seed or reproduce a specific shuffle order; every run is fresh randomness.

Examples

Shuffling a three-column CSV

Input

id,name,active
1,Ada,true

Output

active,id,name
true,1,Ada

The three columns were reordered to active, id, name: note that each column's header and its data cell moved together.

Best Practices & Notes

Best Practices

Use this specifically to test column-order independence, not to anonymize or obscure which value belongs to which field: that's what CSV Value Shuffler is for.
Combine with CSV Row Shuffler if you want both row and column order randomized independently.
Re-run the tool if you want a different random ordering; there's no seed to control the outcome.

Developer Notes

The shuffle draws its random integers from the random category's existing secureRandomInt helper (crypto.getRandomValues with rejection sampling) rather than Math.random, matching how this repo's other randomness-driven tools source entropy.

CSV Column Shuffler Use Cases

Testing that a CSV-consuming system correctly reads by header name rather than assuming column position
Generating varied sample datasets from one source file for demos or testing
Creating a shuffled-column decoy version of a file for a training exercise

Common Mistakes

Expecting this to shuffle individual values: it only moves whole columns, so values within a column stay grouped together.
Using this to anonymize data, when CSV Value Shuffler's per-cell scrambling is the tool actually suited for that.

Tips

If you need only rows shuffled and not columns, use CSV Row Shuffler instead.
Run the shuffle multiple times if you want to compare several different random column orders.

CSV Column Shuffler