Installation#
RAVEN is available as a MATLAB toolbox and as the Python package raven-toolbox. Both build models for the same workflows, some algorithms differ between them (Tuned parameter defaults), and the workflow decides which to use.
Installation of the MATLAB toolbox with the Add-Ons manager, a release download or git, and its upgrade and removal.
pip install --pre raven-toolbox, a development installation, and the upgrade and removal of the package.
Choosing a solver#
Both versions need a linear-programming solver:
Solver |
Licence |
Suited to |
|---|---|---|
Gurobi |
Free academic |
Genome-scale models, ftINIT, gap-filling, and in MATLAB the legacy tINIT method (recommended) |
GLPK |
Open source |
Small and medium models, and getting started |
SCIP |
Open source |
MILP problems (MATLAB); |
In MATLAB, setRavenSolver('gurobi') selects the active solver. In raven-toolbox, the solver is configured through cobrapy (cobra.Configuration).