transparentSampling

A Python toolkit for FRIDA: Data Availability Sampling from FRI, based on the paper "FRIDA: FRI-based Data Availability Sampling" (2024).

Table of Contents

• Features
• Repository Structure
• Prerequisites
• Installation
• Usage
  • Compute FRIDA Scheme
  • Generate LaTeX Tables
  • Generate CSV Data
  • Generate Plots
  • Bin–Ball Sampling
• Data Files
• Citation
• License

Features

• Compute optimal FRIDA parameters (rounds, repetitions) for any data size.
• Multiple schemes: FRIDA, Merkle, Hash, Tensor, TDAS, LT, RS, etc.
• Export performance metrics to CSV.
• Generate LaTeX tables and publication-quality plots.
• Bin–Ball sampling simulation in collectiveBallsInBins/.

Repository Structure

• fri.py Core FRI/FRIDA parameter computation and scheme assembly.
• schemes.py Underlying codes: Reed–Solomon, Hash, Pedersen, etc.
• codes.py Low-level algebraic code definitions.
• Listing4.py Generate summary table in terminal or LaTeX.
• Listing5.py–Listing7.py Export scheme metrics to CSV for different parameters.
• plot_*.py Plot scripts for commitment size, query cost, total cost, encoding, samples.
• collectiveBallsInBins/ Bin–Ball sampling scripts and precomputed plots (*.py, figs/).
• csvdata/ Collected CSV metrics for all schemes.
• figs/ Generated figures (PDF/PNG).
• table/ LaTeX source (table.latex) for summary tables.
• README.md This file.
• LICENSE MIT License.

Prerequisites

• Python ≥3.8
• pip

Installation

git clone https://github.com/your-org/transparentSampling.git
cd transparentSampling
python -m venv venv          # Optional virtual environment
.\venv\Scripts\Activate
pip install pandas numpy matplotlib tabulate

Usage

Compute FRIDA Scheme

from fri import makeFRIScheme
scheme = makeFRIScheme(datasize=32*1024*1024, invrate=4, fsize=128, verbose=True)
print(scheme)

Generate LaTeX Tables

python Listing4.py <datasize_in_MB>         # Default table in text
python Listing4.py <datasize_in_MB> -l      # LaTeX table
# Output LaTeX saved under table/table.latex

Generate CSV Data

python Listing5.py   # Vary datasize
python Listing6.py   # Vary invrate (n/k ratio)
python Listing7.py   # Vary both invrate and k for sample metrics

CSV files are stored in csvdata/.

Generate Plots

python plot_com.py         # Commitment size
python plot_comm_pq.py     # Query cost
python plot_comm_total.py  # Total communication
python plot_encoding.py    # Encoding overhead
python plot_hash_sample.py # Sample count comparison
# Additional scripts with `_kn` suffix for fixed k/n

Figures saved in figs/ or collectiveBallsInBins/figs/.

Bin–Ball Sampling

Navigate to collectiveBallsInBins/ and run:

python plot_samples_to_prob.py   # Plot failure probability vs samples
python plot_to_prob_bin.py       # Bin–probability mapping

See collectiveBallsInBins/figs/ for output.

Data Files

CSV files in csvdata/ are named <scheme>_<metric>.csv, e.g.:

• fri_com.csv Commitment size vs. data dimension
• hash_comm_pq.csv Per-query communication cost
• rs_samples.csv Sample count for RS scheme

Citation

If you use this work, please cite:

J. Doe et al., "FRIDA: FRI-based Data Availability Sampling", XYZ 2024.

License

MIT License – see LICENSE.