The specific parts of these instructions are applicable only to a Debian/Ubuntu Linux.
Nevertheless, these instructions should be adaptable to any modern desktop operating system.
> sudo apt-get install haskell-platform
> sudo cabal update
> cabal update
Note: If you require arbitrary-precision arithmetic using MPFR, you need a modified GHC as described on this page.
> sudo apt-get install libmpfr-dev
> sudo apt-get install libghc-cairo-dev libghc-glade-dev
> git clone https://github.com/michalkonecny/aern.git
> cd aern
> . install-all.sh
If you do not want support for MPFR and/or GTK, use the appropriately named install-*.sh
script instead.
This script invokes cabal-install for each of the AERN packages. See the script code to see in what order the packages are installed.
The first time the script is executed, it is likely to install a large number of dependencies, which can take a long time.
On subsequent executions, the script is likely to give a warning about dangerous reinstalls. This warning can be safely ignored in this case.
Test arithmetic of intervals with Double endpoints:
> ~/.cabal/bin/testAERN-Real-Double
Before running examples that produce plots, make sure your current folder contains the file FnView.glade
. The base folder of the aern
repository contains this file already.
Plot some function enclosures:
> ~/.cabal/bin/plot minmax 10
Compute and plot an enclosure of an ODE IVP with uncertain initial value:
> ~/.cabal/bin/simple-ode ivpSpringMass-uv 10 "PlotGraph[True,False](0,10,-1.5,1.5)" GUI ShrinkWrap 30 200 -10 -10
Compute and plot an enclosure of a hybrid system with Zeno behaviour:
> ~/.cabal/bin/simple-events bouncingBallFloorDropEnergy 4.4 "PlotGraph[True,False,True,False,False](0,5,-0.2,5)" GUI locate evtree 10 10 3 50 0 0
You can also execute various demos using ghci.