How to Implement Operators for Sequence Data Calculation?

[Malibu351]

Hello, and thank you for your work. I've been exploring some operators for sequence data recently, such as calculating the mean of sequence data. However, I noticed that PySR operates on individual data pairs (x1, x2, x3, ...) - (y1) rather than matrices or sequences. Could you advise on how I might configure or modify PySR to achieve the functionality I'm looking for? Thank you in advance for your response.

for example:
I want to use shift(1), when I input series (x1, x2, ... , xn), I can get ( x2, x3, ..., xn, Nan)
or
use mean(x), when I input series (x1, x2, ... , xn) , I can get avg(x), (x1+x2+...+xn)/n

[MilesCranmer]

Related:

• Can I add roll as a custom unary operator #426
• Symbolic integrals with SymbolicRegression.jl #401
• Time-Delayed Variables #275
• Generalization to vector and tensor operators #181

[Malibu351]

Thans a lot, I read #275 before, and I will read others and study how to fix the problem.