2 comments
I don’t understand this post.<p>It seems like the author is studying “disorder” in an array by restricting themselves to looking only at local differences. They make up a lot of reducer functions, many of which relate to each other. But I don’t see the usefulness in playing with them — as none of the individual functions are ever useful. The author even says in their article about Amp that it doesn’t really work.<p>Basically all the functions only observe local behavior, but to measure “disorder” in a sensible way you need to take into account global relationships; the disorder of 2-3-4-5-6-5 is different than that of 2-3-4-5-6-1. There doesn’t seem to be a proposed way around this, but maybe I missed something.