SaatyPy is a Python package for Analytic Hierarchy Process (AHP) and Analytic Network Process (ANP) decision modeling.
It provides robust, flexible, and well-tested tools for multi-criteria decision analysis, supporting both academic and business use cases.

The package is named in honor of Thomas L. Saaty, the originator of AHP and ANP methodologies.
His pioneering work laid the foundation for modern multi-criteria decision making.

• AHP and ANP model builders
• Pairwise comparison matrices
• Consistency checks and error handling
• Matrix limiters and normalization utilities
• Reporting and result export
• Extensible and type-safe API

• saatypy.ahp: Build and solve AHP models (hierarchies, priorities, reporting)
• saatypy.anp: Build and solve ANP models (networks, supermatrices, limiters)
• saatypy.components: Core types, pairwise comparisons, error classes, math utilities
• saatypy.reporting: Generate reports and export results

pip install saatypy

Or clone the repository and install locally:

git clone https://github.com/ehsanAhmadzadeh/saatypy.git
cd saatypy
pip install .

from saatypy.ahp import AHPBuilder

model = (AHPBuilder()
    .add_criteria(["price", "quality", "service"])
    .add_alternatives(["A", "B", "C"])
    .build())

model.set_criteria_weights({"price": 0.5, "quality": 0.3, "service": 0.2})
for crit in ["price", "quality", "service"]:
    model.set_alt_priorities(crit, {
        ("A", "B"): 2.0,
        ("A", "C"): 3.0,
        ("B", "C"): 1.5
    })

priorities, labels = model.alternative_priorities()
print(dict(zip(labels, priorities)))

from saatypy.anp import ANPBuilder
from saatypy.components.pairwise import PairwiseComparison

model = (ANPBuilder()
    .add_cluster("criteria", ["C1", "C2"])
    .add_alternatives(["A", "B"])
    .build())

model.set_cluster_weights({"criteria": 0.6, "Alternatives": 0.4})

block_alt_given_crit = {
    "C1": PairwiseComparison.from_judgments(["A", "B"], {("A", "B"): 4.0}),
    "C2": PairwiseComparison.from_judgments(["A", "B"], {("A", "B"): 3.0/7.0}),
}
model.add_block("Alternatives", "criteria", block_alt_given_crit)
model.add_block_uniform("criteria", "criteria")

priorities, labels = model.alternative_priorities()
print(dict(zip(labels, priorities)))

SaatyPy provides detailed error messages for invalid inputs, missing weights, and matrix inconsistencies:

from saatypy.ahp import AHPBuilder
from saatypy.components.errors import NormalizationError

model = AHPBuilder().add_criteria(["A", "B"]).add_alternatives(["X", "Y"]).build()
try:
    model.set_criteria_weights({"invalid": 1.0})
except NormalizationError as e:
    print("Error:", e)

Extract results directly from your model:

report = model.to_report_data()
print(report["criteria_weights"])
print(report["global_priorities"])
print(report["ranking_str"])

You can save a formatted report for any model using the ReportManager:

from saatypy.reporting.report import ReportManager

manager = ReportManager()  # Default: Markdown format, saves to 'reports/'
report_path = manager.save(model)  # model can be AHPModel or ANPModel
print(f"Report saved to: {report_path}")

# To use plain text format:
from saatypy.reporting.report import PlainRenderer
manager = ReportManager(renderer=PlainRenderer())
report_path = manager.save(model, path="my_report.txt")

• By default, reports are saved in Markdown format to the reports/ directory.
• You can specify a custom path, file name, or renderer.
• The filename will include a timestamp unless you set add_timestamp=False.

• Hierarchical criteria (AHP)
• Custom limiters (ANP)
• Extend with new adapters or reporting formats
• Use math utilities for normalization and consistency checks

• AHP Guide
• ANP Guide
• Usage Guide

• Saaty, T. L. (1980). The Analytic Hierarchy Process. McGraw-Hill.
• Saaty, T. L. (1996). Decision Making with Dependence and Feedback: The Analytic Network Process. RWS Publications.
• Saaty, T. L. (2008). Decision making with the analytic hierarchy process. International Journal of Services Sciences, 1(1), 83–98.
• Saaty, T. L. (2005). Theory and Applications of the Analytic Network Process. RWS Publications.

See CONTRIBUTING.md for guidelines.

See LICENSE.

See CHANGELOG.md.

For more details, see the documentation in the docs/ folder.