A nice demo for the sensitivity of the parameters F
and kappa
is given by
Prof. Trefethen (click
here).
The first parameter set generates a beautiful pattern with "rolls". The parameters for this solution are
F = 0.04
kappa = 0.06
By slightly perturbing F
a completely different solution is found, this time a
"spot" pattern is observed. The parameters for this simulation are
F = 0.025
kappa = 0.06
To reproduce these images, run the command
./run_simulation.py --demo
The Gray-Scott model is implemented in python/gray_scott.py
. A simulation
wrapper around the GrayScott
class is implemented in
python/run_simulation.py
. A simulation can be started with
./run_simulation.py
which integrates the system until time T=3500
using the default parameter
F=0.04
and k=0.06
(kappa). Output is generated every 100
integration
steps. A list of all options can be obtained with
./run_simulation.py --help
Running a simulation for parameter F=0.02
and k=0.06
is achieved by
./run_simulation.py -F 0.02 -k 0.06 --outdir sim_F0.02_k0.06
where in addition the output is written to a directory sim_F0.02_k0.06
in the
current folder. If the --outdir
option is omitted then output is written into
the current directory.
If the option --movie
is specified, the simulation wrapper will generate a
movie from dumped frames. This option requires that ffmpeg
is installed on
the system (ffmpeg download). The default
frame dump frequency is every 100
integration steps. For a high-resolution
(temporal resolution) specify more frequent dumps with the --dump_freq
option.
For example, 10-times more frames are generated with --dump_freq 10
.