Frequency Analysis Tool

Back to Lab Home

About This Tool

This tool allows you to analyze, filter, and cluster numeric frequency data based on percentage difference thresholds. It provides three different algorithms to process your data:

Filter Similar Numbers

Removes numbers that are too close to each other based on percentage difference threshold.

Cluster Similar Numbers

Groups numbers that have at least one neighbor within the threshold, discarding isolated numbers.

Find Largest Similar Group

Finds the largest group where every number differs less than the threshold from all others in the group.

Upload & Configure

No file selected

Upload a text file containing comma-separated frequency values

Filter Similar Numbers

This algorithm removes numbers that are too close to each other based on a percentage difference threshold.

Starting with the lowest number, each subsequent number is kept only if it differs from all previously kept numbers by at least the threshold percentage.

Best for: Creating a set of representative frequencies with guaranteed minimum difference between any two values.

Cluster Similar Numbers

This algorithm identifies clusters of numbers where each number has at least one similar neighbor within the threshold.

Numbers are grouped into clusters where each number differs by less than the threshold from at least one other number in the cluster.

Isolated numbers that don't have any similar neighbor are discarded.

Best for: Identifying groups of related frequencies and eliminating outliers.

Find Largest Similar Group

This algorithm finds the largest possible group where every number differs from all others in the group by less than the threshold.

For each starting number, it builds the largest possible group where every pair of numbers differs by less than the threshold.

Unlike clustering, this ensures complete similarity across all members (not just neighboring pairs).

Best for: Finding a fully cohesive group where every frequency is similar to every other frequency in the group.

%

Minimum percentage difference threshold between numbers (default: 0.025%)

How It Works

Filter Similar Numbers

1

Start with an ordered list of all numbers

2

Add the first number to the filtered set

3

For each subsequent number, check if it differs from all previously kept numbers by at least the threshold

4

If it differs enough, keep it; otherwise, discard it

Cluster Similar Numbers

1

Compare each number with all others to find similar pairs (difference < threshold)

2

Create initial clusters where each number has at least one similar neighbor

3

Merge clusters that have similar members between them

4

Keep only clusters with at least 2 members, discarding isolated numbers

Find Largest Similar Group

1

For each number in the dataset, try to build a group starting from that number

2

Add other numbers to the group only if they differ from all current group members by less than the threshold

3

Verify that every pair of numbers in the resulting group has a difference less than the threshold

4

Keep the largest valid group found across all starting numbers

Privacy and Data Protection

Data Management Information

This tool is designed to protect your privacy and ensure the security of the data you upload.

Automatic File Deletion

  • Results, log, and test files are immediately deleted after download
  • The original file is deleted after you have downloaded all results
  • All files are automatically deleted after 24 hours

Security and Transparency

  • Download buttons become disabled after use for visual feedback
  • No data is shared with external services or third parties
  • All processing takes place on the local server without external dependencies

Liability Disclaimer

This tool is provided "as is" without warranties of any kind, either expressed or implied.

  • The use of this service is at the user's own risk and responsibility
  • We assume no responsibility for the data entered, the accuracy of the results, or any consequences arising from the use of this tool
  • By uploading files to this tool, the user assumes all legal responsibilities related to the content of the files and decisions made based on the results