DINJO 0.1

DINJO Is Not Just An Optimizer is a Python framework designed for the optimization of initial value problems’ parameters.

Lets say you have some ‘experimental’ data of a state variable \(S\) corresponding to the initial value problem

\[ \begin{align}\begin{aligned}d\mathbf{S}/dt &= \mathbf{f}(t, S; \mathbf{p})\\\mathbf{S}(t_0) &= \mathbf{f}_0,\end{aligned}\end{align} \]

where \(\mathbf{p}\) is a list of parameters, \(\mathbf{f}_0\) and \(t_0\) are constants.

If you want to know the optimal value of \(\mathbf{p}\) so that the solution of the initial value problem fits your experimental data, you can use DINJO to get an approximate value of the optimal \(\mathbf{p}\).

Getting Started

Install DINJO using PyPI:

pip install dinjo

Or directly from the latest dev versioN, using source code:

git clone https://github.com/fenfisdi/dinjo
cd dinjo
python setup.py install

Start using DINJO!

The Source Code

dinjo

Package source code

dinjo.model

Define your own initial value problems (IVPs) and solve them using this module.

dinjo.optimizer

Define your IVP optimization problem and solve it using this module.

dinjo.predefined

Some predefined models

dinjo.predefined.epidemiology._seir_model

SEIR initial value problem.

dinjo.predefined.epidemiology._seirv_model

SEIRV initial value problem.

dinjo.predefined.epidemiology._seirv_fixed

SEIR initial value problem.

dinjo.predefined.epidemiology._sir_model

SIR initial value problem.

dinjo.predefined.physics._harmonic_oscillator

Unit mass Harmonic Oscillator initial value problem. See Examples.

Indices and tables