BehaviorSim Documentation
Welcome to the official technical documentation for BehaviorSim—an open-source Python framework and cloud API for generating synthetic sequential behavioral data.
What is BehaviorSim?
BehaviorSim generates sequential data by modeling agents as discrete state machines. At each step, an agent transitions according to stochastic Markov probabilities or rule conditions, emitting continuous and discrete features calibrated for specific domains (finance, healthcare, education, mobile apps).
Who is This Documentation For?
- Researchers: Prototyping sequential algorithms, testing Markov state estimators, and studying latent recovery.
- Machine Learning Engineers: Building feature pipelines, testing recurrent models, and evaluating drift detection without sensitive user data.
- Software Engineers: Stress-testing data ingestion queues, running integration suites, and seeding staging databases.
Quick Installation
pip install behaviorsim==1.0.1Requires Python >= 3.9. Learn more in the Installation Guide.
Key Documentation Areas
Getting Started
Install the package, set up virtual environments, and generate your first simulation trace.
Core Concepts
Understand discrete states, transition matrices, condition rules, and parametric emissions.
Domain Presets
Explore the 4 verified presets: finance, healthcare, education, and mobile application.
Python API & CLI
Method signatures for Simulator, State, FeatureDistribution, and behaviorsim CLI flags.