Mention in paper that marginal likelihoods in the literature may be suspect about kima HOT 2 CLOSED

I wrote this, also trying to respond to #24. Feel free to edit or add something if you want.

Unlike similar open-source packages, kima calculates the fully marginalized likelihood, or
evidence, both for a model with a fixed number Np of Keplerian signals, or marginalising
over Np . For this latter task, Np itself is a free parameter and we sample from its posterior
distribution using the trans-dimensional method proposed by Brewer (2014). Because
kima uses the Diffusive Nested Sampling algorithm, the evidence values are still accurate
when the likelihood function contains phase changes which would make other algorithms
(such as thermodynamic integration) unreliable.

from kima.

I'm happy with it.

from kima.