Skip to content

Latest commit

 

History

History
46 lines (31 loc) · 1.21 KB

File metadata and controls

46 lines (31 loc) · 1.21 KB

Tabular Examples

Differentially private synthetic tabular data with Private Evolution (PE). Each script downloads a public dataset (train/test split + metadata), runs PE, and writes results to results/tabular/<experiment>/:

  • synthetic_tab/ — generated synthetic CSVs
  • checkpoint/ — per-iteration checkpoints (runs resume from here)
  • classifier accuracy (TabClassifier) and Wasserstein-style marginal distance (ComputeWSD) logged to log.txt and CSV

Our code automatically downloads the datasets, which are available at https://github.com/toan-vt/cloud-data-store/tree/main/tabular.

Installation

pip install private-evolution[tabular]

Or, for an editable install from the repo root:

pip install -e ".[tabular]"

Simulated / stress tests

XOR datasets with a configurable number of features (from 1 to 7):

python xor_stress_test.py --num-features 1
python xor_stress_test.py --num-features 2

Structural causal model (SCM) data with a selectable prior function:

python scm.py --prior-function rff   # choices: tree, nn, rff

Real datasets

python artificial_characters.py
python person_activity.py
python adult.py
python breast_cancer.py