Skip to content

Installation

Setting up the CLiP Protocol on your device is simple and quick!

📦 Install via PyPI

Run the following command to install: 

pip install clip-protocol

📎 Note: You can also view the package on PyPI.

⚙️ Usage

Once installed, you can execute the following commands to run the privacy adjustment methods:

Setup

Prepares your dataset for the CLiP workflow. This step formats and validates your input data for the next phases.

setup -d <dataset>
  • <dataset>: Path to the input dataset in .xlsx format.

Mask

Applies personalized local differential privacy to your dataset. This command privatizes the data based on individual privacy budgets.

mask -d <dataset> -o <output>
  • <dataset>: Path to the input dataset you want to privatize.
  • output: Path to where the privatized dataset will be saved.

📎 Note: After masking, a new .csv file will be created containing the privatized data.

The output variable is optional, if it is not needed to save the privatized data you can skip it

Aggregation

Combines the privatized data points into frequency sketches. This command updates the server-side structures needed for final analysis. 

agregate
This command updates the frequency sketches based on the privatized inputs.

Estimation

Estimates the true frequencies from the aggregated privatized data. This command answers frequency queries based on the collected sketches. 

estimate

Clear

Use this command when it is needed to delete all data saved from the previous steps. 

clip_clear

Important Notes

  • ✅ Ensure that dataset paths are correct and accessible.

  • 🛡️ Make sure you have permission to read/write at the specified locations.

  • 📄 The mask step will output a .csv file containing the privatized version of your dataset.