SysIdentPy

Build Dynamic Models

With SysIdentPy

From Classical Approaches to Neural Networks

Get Started
hero

Companies and Universities using SysIdentPy!

SysIdentPy is a simple and easy to learn System Identification library in Python!

shape shape

Model Structure Selection

Use State of the Art techniques to build your models.

Learn More
shape shape

Parameter Estimation

Use recursive methods, adaptive filters and many more.

Learn More
shape shape

Multiple NARMAX Classes

Create Polynomial, Fourier and Neural NARX models.

Learn More

Quick & Easy model building

Build and train NARMAX models easily using intuitive methods! SysIdentPy lets you create and use models of dynamic systems not easily modeled from first principles or specifications..

Try it
about
shape
shape

Model Validation with Essential Features

You can simulate pre-defined models, compare identified models, analyze their properties, and validate them against test datasets..

Try it
about

Robust documentation and examples

There are tons of interesting examples to help you learn SysIdentPy. You can start with our official tutorials right now!

Try it
about
shape
video

Cool facts about SysIdentPy

SysIdentPy empowers anyone to build low-code and robust dynamic models from input and output data.

+12K

Downloads

+80%

coverage

+400

downloads/month

Meet Our Developers

Team
Wilson Rocha

Main Developer and Maintainer

Users sharing their experience

The greatest thing about SysidentPy is that it solves complex problems in a simple and elegant way. Also it has great performance and is very user friendly. We have it now running in production.

shape author
Neylson Crepalde

CDO, A3Data

SysIdentPy is an great to work with time series and dynamic systems, providing native methods and supporting many different estimators from packages like sklearn and Catboost to build different NARMAX models.

shape author
Nath Keles

Data Scientist, RD

In my experience, SysIdentPy is the best python package for System Identification which uses NARMAX models.

shape author
Mehran Shahpari

MSc. Civil - Earthquake engineering | University of Science and Culture, Tehran, Iran

SysidentPy is a high-performance solution that can be used in highly challenging scenarios for non-linear dynamic modeling. At Technium - IA For EveryOne we recommend its use in our projects..

shape author
David Quintão

COO, Technium - AI for EveryOne

Our Recent Blog Posts

blog

Open Source Packages (in portuguese)

Learn More
blog

Nubank Meetup Presentation (in portuguese)

Learn More
blog

GCoM Meetup (in portuguese)

Learn More