My humble contribution to the ongoing quest for mathematicallyoriented sequencers. The idea was based on a paper called "Chaos Melody Theory" (https://www.ziaspace.com/_academic/page148/).
Simple but effective, it's not so easy to find formulas and values that don't spiral upwards or downwards to infinity, but the results are nice sometimes.

Chaos sequencer

@weightless very nice. and i see the iterations of your expr.cell abstraction in use.

@Jona Thanks! I wanted to clean this up for a while and got inspired by your sequencers
Yes that's an interesting way to use the expr.cell, maybe somebody can find a use for this concept too. I really like the idea to be able to address the states of other cells, for example for logics (play this if the last state on cell X was n, stuff like that).
It's all trial an error but the sequences that come out of such systems are very beautiful and complex. 
@weightless if you have a formula that spirals upwards or downwards you can put it into a sinewave function then its always in the range between 1 and 1. thats what we did in the spiro seq and its also implemented in the lissajous seq. i tried to implement your formula and the expr.cell abstraction in our sequencer. nice way to use iterations. this is the result. chaosseqc.zip :
ps. somehow i get only midi output and no audio output from your sequencer, but every object is recognized. 
@Jona Thanks for the tip on using a sine function, pretty great workaround. Here is an updated version with slightly weirder formulas that crossreference each other, wrapped by sin(). There's also a better object to remap the (1 1) values and now you might be hearing audio as well.
I've got to say that it's a lot nicer to be able to see these formulas like in the Lissajous and Spiro sequencers, but this is simple enough that I can make sense of what happens in the patch (so to speak).

@weightless thanks. now i have audio. nice way to use clone for polyphony. i like the melody from your modified formula better than the old one also a nice effect too have different metros (with multiplied values) for each voice.