PAMOLA.CORE

---

Privacy Engineering for Python. Finally.

PAMOLA.CORE is the open-source foundation of the PAMOLA platform — a Python library for privacy-preserving data operations with reproducible pipelines, structured artifacts, and an audit trail per operation.

Developed by Realm Inveo Inc.

---

The Problem

You need to anonymize sensitive data. You've tried:

• ARX: Powerful, but Java, GUI-focused, opaque operations.
• Faker + Presidio + custom scripts: Fragmented, no pipeline, no proof.
• DP libraries: Great math, but narrow scope.

You're still missing:

• Direct operations (mask, generalize, pseudonymize, suppress) — not just "achieve k-anonymity".
• Risk measurement (privacy/fidelity/utility metrics on real outputs).
• Reproducibility (config + metrics + artifacts written to disk per task).

---

The Solution

PAMOLA.CORE: operations-first privacy engineering with a per-operation lifecycle (validate → load → process → save → metrics → visualize → cache).

from pathlib import Path

import pandas as pd

from pamola_core import HashBasedPseudonymizationOperation, FullMaskingOperation
from pamola_core.utils.ops.op_data_source import DataSource
from pamola_core.utils.tasks.task_reporting import TaskReporter
from pamola_core.utils.progress import HierarchicalProgressTracker

df = pd.read_csv("customers.csv")
data_source = DataSource(dataframes={"main": df})
task_dir = Path("./anonymize_customers")
reporter = TaskReporter(task_dir=task_dir, task_name="anonymize")
tracker = HierarchicalProgressTracker(total=2, description="Anonymize")

# 1) Irreversible hash-based pseudonymization of the email column
HashBasedPseudonymizationOperation(
    field_name="email",
    algorithm="sha3_256",
    salt_config={"source": "parameter", "value": "ab" * 32},
    use_pepper=True,
).execute(data_source, task_dir, reporter, tracker)

# 2) Mask the phone column entirely
FullMaskingOperation(
    field_name="phone",
    mask_char="*",
).execute(data_source, task_dir, reporter, tracker)

Output structure (task_dir/):

anonymize_customers/
├── config.json           # Operation configuration (secrets redacted)
├── output/               # Anonymized data (csv/parquet)
├── metrics/              # Privacy & quality metrics (JSON)
├── visualizations/       # Generated charts (PNG)
└── logs/                 # Per-task execution log

Security note (1.0.0.dev3): config.json now redacts sensitive parameters (e.g. AES-256 mapping encryption keys) via OperationConfig.SENSITIVE_KEYS. See CHANGELOG.md for the full release notes.

---

PAMOLA Ecosystem

PAMOLA.CORE is part of a comprehensive privacy engineering stack:

Component	Description	Availability
PAMOLA.CORE	Anonymization, profiling, transformation, metrics, and shared op framework	Open Source (this repo)
PAMOLA.STUDIO	Visual environment for data transformation and privacy management	Commercial
PAMOLA.SYNT	Synthetic data generation, including formal DP-SGD-based generators	Commercial
PAMOLA.BEST	Best-practice policy modules and DP accounting	Commercial
PAMOLA.TEXT	Long text and document anonymization (NLP/LLM-based)	Commercial
PAMOLA.INSIGHT	Agent modules for LLM integration	Commercial

Scope of CORE: This package provides anonymization and pseudonymization primitives, classical privacy metrics, and reproducibility plumbing. It does not implement formal differential-privacy synthetic-data generation — that lives in PAMOLA.SYNT / PAMOLA.BEST.

---

What's In CORE

All classes below are exported from the top-level pamola_core package.

Category	Operations
Anonymization — Masking	FullMaskingOperation, PartialMaskingOperation
Anonymization — Generalization	CategoricalGeneralizationOperation, NumericGeneralizationOperation, DateTimeGeneralizationOperation
Anonymization — Suppression	AttributeSuppressionOperation, CellSuppressionOperation, RecordSuppressionOperation
Anonymization — Pseudonymization	HashBasedPseudonymizationOperation (SHA3-256/512 + salt + pepper), ConsistentMappingPseudonymizationOperation (AES-256-GCM reversible mapping)
Anonymization — Noise	UniformNumericNoiseOperation, UniformTemporalNoiseOperation
Profiling — Field analyzers	CategoricalOperation, CorrelationOperation, CorrelationMatrixOperation, CurrencyOperation, DateOperation, EmailOperation, GroupAnalyzerOperation, IdentityAnalysisOperation, MVFOperation, NumericOperation, PhoneOperation, TextSemanticCategorizerOperation
Profiling — Dataset-level	KAnonymityProfilerOperation, DataAttributeProfilerOperation
Transformation	AddOrModifyFieldsOperation, RemoveFieldsOperation, CleanInvalidValuesOperation, ImputeMissingValuesOperation, AggregateRecordsOperation, MergeDatasetsOperation, SplitByIDValuesOperation, SplitFieldsOperation
Fake Data	FakeNameOperation, FakeEmailOperation, FakePhoneOperation, FakeOrganizationOperation
Metrics	FidelityOperation (KS, KL-divergence), PrivacyMetricOperation (DCR, NNDR, uniqueness, k-anonymity, l-diversity), UtilityMetricOperation (classification, regression)
Analysis helpers	analyze_dataset_summary, analyze_descriptive_stats, analyze_correlation, visualize_distribution_df, calculate_full_risk

Note on attack simulation: Internal modules under pamola_core/attacks/ (linkage, attribute inference, membership inference, DCR/NNDR helpers) are used by PrivacyMetricOperation, but there are no public, registered Attack*Operation classes in CORE 1.0.0.dev3. Future releases may expose them.

---

Pseudonymization Spotlight (1.0.0.dev3)

The 1.0.0.dev3 release hardened the pseudonymization stack:

from pamola_core import ConsistentMappingPseudonymizationOperation

op = ConsistentMappingPseudonymizationOperation(
    field_name="customer_id",
    mapping_encryption_key="ab" * 32,         # 256-bit hex key
    pseudonym_type="uuid",                    # or "sequential" / "random_string"
    mapping_format="csv",                     # encrypted at rest with AES-256-GCM
    persist_frequency=1000,
)

# Outputs:
#  - {task_dir}/output/         anonymized data
#  - {task_dir}/output/         encrypted mapping file (re-identification key)
#  - {task_dir}/metrics/        operation metrics
#  - {task_dir}/config.json     mapping_encryption_key is replaced with "*REDACTED*"

Highlights:

• AES-256-GCM mapping encryption keys are never persisted to disk. A new OperationConfig.SENSITIVE_KEYS declaration + to_safe_dict() redacts secrets before any save_config() call.
• Hash-based op rejects weak salts (all-zero or empty) when use_pepper=False.
• Per-run session id invalidates stale disk cache when use_pepper=True, so previous-run pseudonyms cannot be served back.
• Compound identifiers, ENRICH/REPLACE modes, reverse mapping, and Dask pickle safety are covered by 41 dedicated tests.

---

Metrics

Metric operations write JSON artifacts under {task_dir}/metrics/:

from pamola_core import FidelityOperation, PrivacyMetricOperation

FidelityOperation(
    fidelity_metrics=["ks", "kl"],
    columns=["age", "income"],
).execute(data_source, task_dir, reporter, tracker)

PrivacyMetricOperation(
    privacy_metrics=["dcr", "nndr", "uniqueness"],
    quasi_identifiers=["age", "gender", "zipcode"],
).execute(data_source, task_dir, reporter, tracker)

For dataset-level utility scoring (classification / regression downstream models) use UtilityMetricOperation.

---

Installation

From PyPI:

pip install --pre pamola-core

The --pre flag is required. Every 1.0.0 release so far is a pre-release, and pip skips pre-releases by default — so a plain pip install pamola-core will not give you this library. You can also pin the exact version instead: pip install pamola-core==1.0.0.dev4. The flag stops being necessary once 1.0.0 final ships.

From source:

git clone https://github.com/DGT-Network/PAMOLA.git
cd PAMOLA
pip install -e .

Test extras:

pip install -e ".[test]"   # adds pytest, pytest-cov

Heads-up: All scientific dependencies (numpy, pandas, scikit-learn, scipy, dask, spacy, faker, cryptography, etc.) are pinned in the main [project.dependencies] table — no separate [fast]/[ner]/[dp] extras in this release.

As of 1.0.0.dev4, torch and sdv are no longer installed: nothing in the library imported them. If you relied on them being pulled in as a side effect, install them explicitly. Splitting the remaining heavy dependencies (spacy, nltk, faiss, dask, the plotting stack) into optional extras is planned for a later release.

---

Supported Python Versions

PAMOLA.CORE supports Python 3.10, 3.11, and 3.12 (requires-python = ">=3.10,<3.13").

Python Version	Supported
3.10	✅
3.11	✅
3.12	✅
3.9 and below	❌
3.13 and above	❌ (blocked by third-party dependencies)

---

Core Dependencies

A non-exhaustive view of the heaviest third-party packages (full list in pyproject.toml):

Package	Pin	Purpose
numpy	1.26.4	Numerical computation across metrics, attacks, statistical analysis
pandas	2.2.2	The DataFrame container for every CORE operation
scikit-learn	1.7.2	Classification/regression metrics, nearest-neighbor distance, model-based utility
scipy	1.15.3	Statistical tests (KS, KL divergence) used by FidelityOperation
cryptography	46.0.3	AES-256-GCM mapping encryption for ConsistentMappingPseudonymizationOperation
dask[complete]	2025.11.0	Optional out-of-core / distributed execution path
pyarrow	14.0.2	Parquet I/O
typer	0.24.1	CLI entry point

---

Versioning

PAMOLA.CORE follows Semantic Versioning and PEP 440.

import pamola_core
print(pamola_core.__version__)   # e.g. "1.0.0.dev3"

Phase	Version	Branch	Tag	Install
Dev (current)	1.0.0.dev3	develop	v1.0.0.dev3	pip install pamola-core==1.0.0.dev3
Stable (planned)	1.0.0	main	v1.0.0	pip install pamola-core

• Source of truth: pyproject.toml → version
• Changelog: CHANGELOG.md
• CI/CD: GitHub Actions — lint (ruff), test (3.10/3.11/3.12, pytest), build (sdist+wheel), PyPI publish on tag v*
• Release rules: Dev tags (v*dev*) must be on develop; stable tags on main.

---

CLI

The pamola-core console script is installed automatically:

pamola-core --version
pamola-core list-ops                              # discover registered operations
pamola-core run --task task.json                  # run a task definition
pamola-core run --op FullMaskingOperation --config config.json --input data.csv
pamola-core schema FullMaskingOperation           # show parameter schema
pamola-core validate-config --config config.json  # validate a config file

Run pamola-core --help for the full command list.

---

Examples

Hands-on notebooks live under examples/:

• examples/anonymization/pseudonymization/ — simple + advanced for hash-based and consistent-mapping pseudonymization
• examples/anonymization/ — masking, generalization, noise, suppression
• examples/profiling/ — field-level and dataset-level profilers
• examples/transformations/ — merge, split, aggregate, clean, impute
• examples/fake_data/ — synthetic identity, email, phone, organization
• examples/metrics/ — fidelity, privacy, utility metrics
• examples/data_examples/sample.csv — non-PII synthetic sample used by the notebooks

No real personal data is included in this repository. All sample records are artificially generated.

---

Philosophy

• Operations-first: Direct transforms with a well-defined 7-step lifecycle, not constraint optimization.
• Measure everything: Privacy, fidelity, and utility metrics persisted as JSON artifacts.
• Reproducibility by default: Each operation writes a config.json (with secrets redacted) alongside its output.
• Secret hygiene: OperationConfig.SENSITIVE_KEYS provides a single place to declare parameters that must never reach disk.

---

API Documentation

The project uses Sphinx to generate API reference documentation from Python docstrings.

cd docs
make html        # output: docs/_build/html/index.html

---

Documentation

Resource	Link
PET Knowledge Base	realmdata.io/kb
Technical Documentation	docs/en/index.md
Glossary	realmdata.io/glossary
Examples	examples/
Changelog	CHANGELOG.md

---

Use Cases

• Data Engineering: Prepare privacy-safe datasets for ML training.
• Healthcare: HIPAA-oriented de-identification workflows (Safe Harbor support).
• Finance: Privacy engineering aligned with PCI/GDPR considerations.
• Compliance: Audit-ready evidence with structured per-operation artifacts.
• Data Sharing: Risk-assessed data exchange between organizations.

---

Regulatory Context

PAMOLA.CORE provides technical building blocks for privacy compliance programs:

Regulation	Relevant Capabilities
GDPR	Pseudonymization (reversible / irreversible), data minimization (Art. 25, 32)
HIPAA	Safe Harbor de-identification support
CCPA/CPRA	Data suppression, masking, anonymization workflows

Important: PAMOLA.CORE provides technical capabilities only. Legal compliance requires organizational policies, procedures, and legal guidance beyond software tools.

---

Contributing

git clone https://github.com/DGT-Network/PAMOLA.git
cd PAMOLA
pip install -e ".[test]"
pytest tests/ -v

See CONTRIBUTING.md for guidelines.

---

Ownership & Licensing

PAMOLA.CORE is developed and owned exclusively by Realm Inveo Inc.

This repository is hosted under the DGT-Network GitHub organization, which provides shared development infrastructure for Realm Inveo projects. DGT-Network does not claim ownership of this intellectual property. All IP rights belong exclusively to Realm Inveo Inc.

License: BSD 3-Clause — see LICENSE.

---

Contact

Purpose	Contact
General inquiries	contact@realmdata.io
Commercial / Sales	sales@realmdata.io
Due diligence / Legal	legal@realmdata.io
Website	realmdata.io

---

_{Built by Realm Inveo Inc.}