The goal of this PR is to implement the halofit recipy, in order to have a nonlinear power spectrum built on top of the linear prediciton of Bolt.jl.
In order to achieve that goal, I have:
added some functions to utils.jl, required to compute the variance of (filtered) fluctations
added a new module, nonlinear.jl, that computes the halofit power spectrum
added a script which shows how to use my implementation
I haven't yet fully inspected the performance side, but I suppose that the most time consuming part is the evaluation of the scale at which the variance is equal to one.
The goal of this PR is to implement the halofit recipy, in order to have a nonlinear power spectrum built on top of the linear prediciton of Bolt.jl.
In order to achieve that goal, I have:
utils.jl, required to compute the variance of (filtered) fluctationsnonlinear.jl, that computes the halofit power spectrumI haven't yet fully inspected the performance side, but I suppose that the most time consuming part is the evaluation of the scale at which the variance is equal to one.