Code to accompany OpenDP Library 0.11 Blog

0.11.py

import polars as pl
import opendp.prelude as dp
dp.enable_features("contrib")

# set up your analysis
context = dp.Context.compositor(
    data=pl.scan_csv("pet_species.csv"),
    privacy_unit=dp.unit_of(contributions=1),
    privacy_loss=dp.loss_of(epsilon=1., delta=1e-7),
    split_evenly_over=2)

# release counts of the number of pets per-species
query = context.query().group_by("species").agg(
    pl.len().dp.noise()
)

# view calibrated scale parameters per-query, accuracy estimate and censoring threshold
print(query.summarize(alpha=0.05))
# ┌────────┬───────────┬─────────────────┬───────┬──────────┬───────────┐
# │ column ┆ aggregate ┆ distribution    ┆ scale ┆ accuracy ┆ threshold │
# │ ---    ┆ ---       ┆ ---             ┆ ---   ┆ ---      ┆ ---       │
# │ str    ┆ str       ┆ str             ┆ f64   ┆ f64      ┆ u32       │
# ╞════════╪═══════════╪═════════════════╪═══════╪══════════╪═══════════╡
# │ len    ┆ Len       ┆ Integer Laplace ┆ 2.0   ┆ 6.429605 ┆ 33        │
# └────────┴───────────┴─────────────────┴───────┴──────────┴───────────┘

print(query.release().collect())
# ┌─────────┬─────┐
# │ species ┆ len │
# │ ---     ┆ --- │
# │ str     ┆ u32 │
# ╞═════════╪═════╡
# │ cat     ┆ 112 │
# │ dog     ┆ 142 │
# │ mouse   ┆  45 │
# └─────────┴─────┘