A Code Generator for ODE-Based Models

Provides an R6 class and several utility methods to facilitate the implementation of models based on ordinary differential equations. The heart of the package is a code generator that creates compiled 'Fortran' (or 'R') code which can be passed to a numerical solver. There is direct support for solvers contained in packages 'deSolve' and 'rootSolve'.


rodeo is an add-on package for the R statistical computing software. It facilitates the implementation of mechanistic models built on ordinary differential equations (ODE).

  • rodeo imposes a well-known standard notation for simultaneous ODE based on the stoichiometry matrix. This notation is widely used in chemical and wastewater engineering or biogeoscience.

  • rodeo separates the conceptual model (the set of ODE) from its implementation (computer program) by means of automatic code generation. This makes models more easy (1) to understand by non-programmers, (2) to run on different platforms/environments, (3) to develop or modify, and (4) to 'archive' in times of rapidly changing software.

  • The generated code (Fortran, R) can be used with existing suites of numerical solvers, namely the R-packages deSolve for dynamic simulation and rootSolve for steady-state estimation. The use of compiled Fortran speeds up numerical integration significantly compared to a purely R-based model implementation (often by a factor between 2 and 100).

  • rodeo has with built-in support for 1-D partial differential equations (PDE) being tackled with the numerical method-of-lines.

The package currently provides a single R6 class named just like the package (rodeo). It contains several class- and non-class methods to

  • import a conceptual model from tabular text files (or spreadsheets).
  • generate source code to be passed to an ODE-solver.
  • visualize and export basic information about a model, e.g. for documentation purposes.
  • generate a stoichiometry matrix from a set of reaction equations.
  • validate a stoichiometry matrix (mass balance check).

The package vignette provides further details.

A current installation of R is required. In order to work with rodeo in high-performance mode, one needs the tools to compile Fortran 2008 code (compiler, make utilities, etc.). These tools are typically available by default on Linux systems. Users of other systems should have a look at the Rtools.

The development version of the package can be installed directly from the github repository as shown below. For this, the devtools package is required.

library("devtools")
install_github("dkneis/rodeo")
library("rodeo")

The release version should be available on CRAN and can be installed as usual (see below). Note that the release version may lag behind the development version in terms of functionality.

# Install and load release version
install.packages("rodeo")
library("rodeo")

News

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.

install.packages("rodeo")

0.7.1 by David Kneis, 5 months ago


https://github.com/dkneis/rodeo


Browse source code at https://github.com/cran/rodeo


Authors: David Kneis <david.kneis@tu-dresden.de>


Documentation:   PDF Manual  


GPL (>= 2) license


Imports R6, deSolve

Suggests knitr, rmarkdown, xtable, rootSolve

System requirements: The tools to run 'R CMD SHLIB' on 'Fortran' code. The used 'Fortran' compiler must support pointer initialization which is a feature of the 2008 standard.


See at CRAN