Added Cash-C likelihood

[itrharrison]

I have implemented a Cash-C likelihood for use with e.g. the binned cluster counts.

I assumed the data was stored in a particular form:

        data = np.loadtxt(self.datapath, unpack=False)
        N = data[:, -1] # assume data stored like column_stack([z, q, N])
        x = data[:, :-1]

but that might not make sense in practise!

Merging will close #54 .

[itrharrison]

I should also say that the fiducial likelihood values in the test came from using @eunseongleee 's implementation of the Cash likelihood on the binned_clusters branch, so hopefully my implementation is correct.

[eunseongleee]

Looks all good to me!