Normalization in projected correlation function

[Felipe Gómez]

Hi!

I am triying to calculate the projected correlation function w_p of dark mater halos and comparing with previous results (blue line):

https://github.com/flgomezc/correlationz3/blob/master/programs/testing%20astroml-Copy1.ipynb

When I use the whole halo catalog, they are in a cubic box of 250 Mpc h-1 side length. ( green line at the bottom).

When I use parts of the halo catalog, they are cubic boxes of 80 Mpc h-1 side length. ( red line ).

I try to guess the difference factor, for the green line was 250 (as the box size), and for the others was 80 (again, as the box size).

Why appears this factor with the size of the box?

Is it OK to multiply by the boxsize to get the right normalization?

Thanks for your help.

Felipe

Pd:

AstroML is awesome.