About

Assimulo is a Cython/Python based simulation package that allows for simulation of both ordinary differential equations of the form $f(t,y)$ (explicit problems) and differential algebraic equations of the form $f(t, y, y_d)$ (implicit problems). Assimulo currently supports Explicit Euler, adaptive Runge-Kutta of order 4 and Runge-Kutta of order 4. Assimulo also wraps the popular SUNDIALS solvers CVode (for explicit problems) and IDA (for implicit problems). Hairer's codes Radau5, Rodas and Dopri5 is also available.

CVode and IDA supports discontinuous systems of the form $f(t, y, s_w)$ and $f(t, y, y_d, s_w)$ with event(root)-functions defined as $g(t, y, s_w)$ and $g(t, y, y_d, s_w)$ where $s_w$ should be fixed during the integration.

The package comes with a Problem specifications class and subclasses corresponding to different types of problems. Choices are Implicit_Problem and Explicit_Problem. To define and solve a problem, first import the solver of choice and the appropriate Implicit/Explicit class. Furthermore, define your function $f$ and the initial conditions to pass to the problem class constructor. Then, create your solver, set the attributes (method, absolute/relative tolerance etc.) and use the simulate method to simulate.

Sundials Compliance

Current Assimulo development aims for compliancy with Sundials v2.7.0.

Some optional features in `Assimulo>=3.4.1`` are built on the modified Sundials version that can be accessed here.

Installation

The recommended way to install Assimulo is through CONDA.

conda install conda-forge::assimulo

For installing from source, see the INSTALL file.

Note that currently the releases on PYPI is outdated.

Contributing

For information about contributing, see https://github.com/modelon-community/contributing.

Contact

For contact, either via email or the easiest is immediately by filing an issue or pull request here on GitHub. Email: christian.winther at modelon.com.