How about when the probability of a reaction is randomized at each pixel and at each step. The value is uniformly distributed between the p_reaction parameter and 1.0.
Case 10 has the most obvious asymmetry. Let’s animate it.
Coooool.
Also, I hate the way I specify the range of values I want to explore for a given parameter. I have to provide the center number and the size of the window. That means if I wanna look at feed rates between 0.7 and 1.0 I need to specify 0.85 and 0.3.
cases = get_cases(base_case, 'kill_rate', (0.045, 0.6), 'p_reaction', (0.6, 1.0), 6)
Now I can just specify the range I want to look at for each parameter.
Lol whoops, all that and I go and make a mistake setting my ranges. Hahah, classic.
These look cool but really, is it any different from what I’d get without the variation in probability? Maybe if they run longer, hmm…
Note to self: here’s a shell command for duplicating a file multiple times… like duplicating the last frame of an animation:
for i in {81..99}; do cp output/frame_0080.png "output/frame_00$i.png"; done
Here’s the output for case 27. I do like the shapes that form in the center. Here’s that last frame:
Asymmetrical.
Now, how does that compare with something similar without the probability parameter? Here’s case 33 with a probability of 1.0.
Last frame:
Symmetrical.
Nice. Just to double check, case 34.
I mean, I could probably just make the seed asymmetrical and get the same effect, right?
Tried flipping the image horizontally, the image that I generated with the probability parameter. Just to see how different it is … or isn’t.
It’s definitely not symmetrical but it’s interesting how similar it is. Would be neat to measure the symmetry and assign a score. I wonder what the sweet spot is. Completely symmetrical is boring. Completely asymmetrical is boring. Okay, no, all of them can be interesting.