Check the method to obtain data vectors from simulations

[anicola]

We need to check the methodology used to obtain sample-variance reduced data vectors from the simulations.

[damonge]

Besides the method itself, we should also check that the generated data are stable wrt. some of the choices in that method. E.g. we should check:

• That the sampling in z we are using is good enough
• That the choice of interpolation scheme in z and k is not relevant.
• That changing the fitting function used for b(k) is inot relevant.
• That the width of the savgol filter used to smooth-out the small scales is not relevant.

Tagging @carlosggarcia , who expressed interest in this.