Goal : To extract more information from small scales(~50Mpc)
Method : Train some generated simulations with Gaussian Process and using MCMC to recover test simulation parameters.
Result : Using 5 kinds of Correlation functions can predicted with few % error at small scales, improving cosmological parameters precision increase with 15 ~ 25%
Projected : ω_p(r_p)
Monopole : ξ_0(r)
Quadrupole : ξ_2(r)
Underdensity probability : P_U(r)
Marked correlation function : M(r)
Our progress
Considering all possible shapes in 4pcf on our code and working on generalizing issues.