Open Models. Trusted Intelligence. Shared Progress.
- 100+ Production-Ready Models - ARIMA, GARCH, VAR, Kalman Filters, Random Forest, XGBoost, and more
- Built for Public Data - Native integration with Census, BLS, FRED, CDC, HUD via KRL Data Connectors
- Research-Grade Quality - Peer-reviewed algorithms with comprehensive validation and testing
- Production Ready - Battle-tested in real-world policy analysis and forecasting applications
- Fully Documented - Complete API reference, tutorials, and mathematical formulations
- Open Source - Apache 2.0 License, community-driven development
KRL Model Zoo™ is an open-source library of socioeconomic and econometric models designed for researchers, policymakers, analysts, and community organizations seeking to turn data into actionable intelligence.
It provides modular, production-grade tools for forecasting, regional analysis, anomaly detection, and policy evaluation — all engineered with transparency, reproducibility, and accessibility in mind.
Developed by KR-Labs, the Model Zoo bridges statistical rigor with public-interest purpose, transforming raw data into real-world insight.
The Model Zoo is designed to work seamlessly with KRL Data Connectors, our companion library for accessing public datasets. Together, they provide an end-to-end workflow:
Data → Analysis → Insight
- KRL Data Connectors pulls data from Census, BLS, FRED, CDC, HUD, and 20+ other federal sources
- KRL Model Zoo applies statistical models and analytical frameworks
- Results drive evidence-based decisions and policy evaluation
We believe data science should serve humanity.
The KRL Model Zoo exists to make responsible, interpretable, and replicable analytics accessible to everyone shaping economic, social, or community outcomes.
Our models support research, governance, and education by standardizing open methods for understanding change.
- Unified Framework: Econometric, causal, ML, and Bayesian models under one architecture.
- Transparency by Design: Built-in provenance tracking, version control, and reproducibility.
- Seamless Data Integration: Native compatibility with KRL Data Connectors for federal datasets.
- Community-Driven: Collaborative ecosystem for model sharing, validation, and continuous improvement.
- Policy-Ready Tools: Designed for public datasets like Census, BLS, FRED, CDC, and HUD.
- Educational Utility: Ideal for universities, research institutes, and civic technology programs.
| Domain | Example Models | Use Case |
|---|---|---|
| Time Series Analysis | ARIMA, SARIMA, GARCH, Kalman Filters | Forecasting unemployment, prices, or economic volatility |
| Econometrics | VAR, Cointegration, Structural Breaks | Studying policy impacts and long-term relationships |
| Regional Analysis | Location Quotient, Shift-Share | Understanding regional economic specialization |
| Anomaly Detection | STL Decomposition, Isolation Forest | Identifying shocks and unusual patterns in public data |
| Extensions | Causal Inference, Bayesian Hierarchies, Network Models | Enterprise Applications |
The Model Zoo powers work that matters:
- Labor & Employment: Forecasting job trends, analyzing workforce shifts, and tracking equity gaps.
- Housing & Urban Development: Modeling affordability, detecting displacement, and identifying price volatility.
- Income & Inequality: Measuring economic disparity, mobility, and opportunity over time.
- Public Health: Linking health indicators with economic and environmental conditions.
- Regional Development: Assessing industrial strengths, resilience, and competitiveness.
Each model is field-tested, policy-relevant, and community-accessible.
Install from PyPI (recommended):
pip install krl-model-zooInstall with optional dependencies:
# For development
pip install krl-model-zoo[dev]
# For testing
pip install krl-model-zoo[test]
# For documentation
pip install krl-model-zoo[docs]
# All optional dependencies
pip install krl-model-zoo[all]Install from source:
git clone https://github.com/KR-Labs/krl-model-zoo.git
cd krl-model-zoo
pip install -e .Complete ecosystem setup:
# Install Model Zoo + Data Connectors for seamless data access
pip install krl-model-zoo krl-data-connectors# Step 1: Fetch data using KRL Data Connectors
from krl_data_connectors import BLSConnector, CensusConnector
# Get unemployment data from BLS
bls = BLSConnector()
unemployment_data = bls.get_series('LNS14000000', start_year=2015, end_year=2024)
# Get regional employment from Census
census = CensusConnector()
regional_data = census.get_cbp_data(year=2023, geography='county', state='06')
# Step 2: Analyze with Model Zoo
from krl_models.econometric import SARIMAModel
from krl_models import LocationQuotientModel
# Forecast unemployment trends
model = SARIMAModel(
data=unemployment_data,
params={'order': (1, 1, 1), 'seasonal_order': (0, 0, 0, 0)},
meta={'name': 'unemployment_forecast', 'version': '1.0.0'}
)
result = model.fit()
forecast = model.predict(steps=12)
# Analyze regional specialization
lq = LocationQuotientModel(
data=regional_data,
params={
'region_col': 'county',
'industry_col': 'naics',
'employment_col': 'emp'
},
meta={'name': 'regional_lq', 'version': '1.0.0'}
)
lq_result = lq.fit()
# Step 3: Export results for reporting or visualization
forecast.to_csv('unemployment_forecast.csv')
lq_result.to_excel('regional_analysis.xlsx')from krl_models.econometric import SARIMAModel
from krl_models import STLAnomalyModel
import pandas as pd
# Use your own data
data = pd.read_csv('your_data.csv')
# Forecast with SARIMA
model = SARIMAModel(
data=data,
params={'order': (1, 1, 1), 'seasonal_order': (1, 1, 1, 12)},
meta={'name': 'my_forecast', 'version': '1.0.0'}
)
result = model.fit()
forecast = model.predict(steps=12)
# Detect anomalies
anomaly_model = STLAnomalyModel(
data=data,
params={
'time_col': 'date',
'value_col': 'metric',
'seasonal_period': 12,
'threshold': 3.0
},
meta={'name': 'anomaly_detection', 'version': '1.0.0'}
)
anomalies = anomaly_model.fit()Visit /examples/notebooks/ for guided walkthroughs:
- End-to-End Workflows: Fetching data with Connectors + analyzing with Model Zoo
- Forecasting labor market trends with BLS data
- Regional specialization analysis using Census CBP data
- Housing market volatility with HUD Fair Market Rent data
- Anomaly detection in CDC health indicators
- Multi-source integration: combining FRED, BLS, and Census data
KR-Labs thrives on open collaboration and shared intelligence.
You can:
- Contribute new models, tutorials, or datasets
- Report issues or propose features
- Share research Mapplications and case studies
- Join our growing community of open-data practitioners
See our Contributing Guide for details.
Join the discussion at GitHub Discussions.
The Model Zoo evolves through the KR-Labs Gate Framework, ensuring transparent, high-quality development:
| Gate | Description | Status |
|---|---|---|
| Gate 1 – Foundation | Time series & econometric core | Open-Source |
| Gate 2 – Domain Models | Regional & anomaly detection | Open-Source |
| Gate 3 – Ensembles | Meta-models and hybrid systems | Enterprise App |
| Gate 4 – Advanced Extensions | Causal inference, Bayesian, network models | Enterprise App |
Each Gate reflects a maturity milestone balancing innovation, stability, and community feedback.
The Model Zoo is part of a broader open-source intelligence platform:
| Repository | Purpose | Status |
|---|---|---|
| krl-model-zoo | Statistical models & analytical frameworks | Production |
| krl-data-connectors | Unified API for 20+ federal data sources | Production |
Traditional Approach:
- Manually download CSVs from multiple government websites
- Write custom parsers for each data format
- Clean and normalize data structures
- Finally begin analysis
KR-Labs Approach:
pip install krl-data-connectors krl-model-zoo- Write analysis code immediately
- Results in minutes, not days
Key Benefits:
- Single API for Census, BLS, FRED, CDC, HUD, and more
- Pre-validated data compatible with Model Zoo methods
- Reproducible workflows from data fetch to final analysis
- Version control for both data retrieval and model parameters
- Community support for common use cases and patterns
📚 Full Documentation on ReadTheDocs (Coming Soon)
- Quickstart Guide – Get started in 5 minutes
- User Guide – Comprehensive usage documentation
- API Reference – Complete API documentation
- Contributing Guide – Join our open-source community
- Development Guide – Architecture and development practices
- Testing Guide – Quality assurance and test framework
- Python Examples – Standalone Python scripts demonstrating model usage
- Jupyter Notebooks – Interactive tutorials with real datasets
- KRL Data Connectors – Integrate with 20+ federal data sources
- Software: Apache 2.0 License – Free for academic and commercial use
- Documentation: CC-BY-SA-4.0
KR-Labs™ and KRL Model Zoo™ are trademarks of Quipu Research Labs, LLC, a subsidiary of Sundiata Giddasira, Inc.
If you use the KRL Model Zoo in research or analysis, please cite:
@software{krl-model-zoo,
author = {Deloatch, Brandon C.},
title = {KRL Model Zoo: Open-Source Socioeconomic Modeling Framework},
Year = {2025},
publisher = {KR-Labs},
version = {1.0.0},
url = {https://github.com/KR-Labs/krl-model-zoo}
}- Website: krlabs.dev
- Email: [email protected]
- GitHub Discussions: Join the conversation
© 2025 KR-Labs. All rights reserved.
KR-Labs™ and KRL Model Zoo™ are trademarks of Quipu Research Labs, LLC, a subsidiary of Sudiata Giddasira, Inc.
Software License: Apache 2.0 – Free for commercial and academic use
Documentation License: CC-BY-SA-4.0
