“A simple developmental model recapitulates complex insect wing venation patterns” Jordan Hoffman, Seth Donaughe https://doi.org/10.1073/pnas.1721248115

Inspired by the paper above, I started experimenting with generating patterns of veins as on insect wings. For now I’m just exploring reaction-diffusion using code from Doormatty on Reddit.

https://www.reddit.com/r/Python/comments/og8vh6/comment/h4k2408/

Here are some screenshots of outputs generated by varying feed and kill rates.

This image looks closer to the shapes I’m gonna want for the primary veins. Longer veins that branch out to fill the space. I can change my parameters to match and then explore that neighbourhood of the parameter space to find exactly what I’m looking for.

Goal

#

For now I want to constrain the growth to a long and narrow shape with the seed at one end. I’m hoping to find parameters which give me the secondary veins. Ideally, the new vein will split out from the most recent vein in a satisfying way. (-:

The Work

#

That wasn’t too hard, actually. But the outputs are mostly pretty boring. Mostly variations of this:

The most interesting one had some variation vertically, too:

Feed rate: 0.021 Kill rate: 0.048

I just made it easier to define the cases to explore. I specify the target values for feed and kill and the size of the window for each and then the number of cases.

Some of the more promising results from the last batch:

Digging in a little deeper:

Maybe the best of the next set:

Not getting what I’m looking for. Maybe changing the height of the surface or the diffuse rates?

Was gonna write a script to view the images in a grid but turns out ImageMagick has me covered.

montage *.png -background grey -tile 8x -geometry +2+2 ../montage.png