PySTRA: structural reliability in Python#
PySTRA implements established and selected modern structural reliability methods, with practical tools for code calibration and structural assessment. Define your engineering model in Python, quantify its reliability, and inspect the assumptions and diagnostics behind the result.
These pages describe PySTRA 2.0.0. Follow Installation to install it. What’s new in PySTRA 2.0 summarizes the release, and Migrating from 1.x to 2.0 shows how to update 1.x code; the 1.x documentation remains available.
Run a reliability analysis
Build a two-variable model and check FORM against an analytic answer.
Calibrate code factors
Compare candidate factors using normalized reliability and verify the resulting designs.
Compare assessment scenarios
Evaluate how specified resistance and load assumptions change reliability.
Estimate very small probabilities
Tail-accurate transformations and log-space estimators stay finite far beyond where floating-point probabilities round off.
Find the right material#
Get started takes you from installation to a checked first result.
What’s new in PySTRA 2.0 summarizes what 2.0 adds and changes.
User guide helps you choose methods, build models and interpret results.
Examples contains worked tutorials and published benchmark problems.
API reference describes public objects, parameters, results and extension details.
Theory explains the formulations, assumptions and notation.
For method selection, begin with Choosing a reliability method. For unexpected results, see Troubleshooting reliability analyses. To reproduce a paper example, browse the Benchmark catalogue.
Project and community#
PySTRA is GPL-3.0-or-later. Its numerical methods build on the work cited in References, including the original framework [Hackl2013]. See Citing PySTRA and its methods for software citation and method attribution. Report questions and reproducible problems through the issue tracker.