Bounding the output function from PySR

[Khansolo9]

Although I am successfully getting a function for our data - as expected - is there any way I could get an inequality for our data as an output of PySR, as well? Meaning, is there anyway I could bound the data …? E.g. if expected value calculated is <X, then satisfied, otherwise, not.

Best,
Khansolo9

P.S. thoroughly enjoy using this software!

[MilesCranmer]

Thanks!

Like this?

model = PySRRegressor(loss="loss(x, y) = (x <= y) ? ((x-y)^2) : Inf")

[MilesCranmer]

Although I might choose a large number rather than infinity. i.e.,

model = PySRRegressor(loss="loss(x, y) = (x <= y) ? ((x-y)^2) : 100000")

that makes the constraint softer and easier to explore.