Open RobinHankin opened 6 months ago
However this suggestion causes ugly equations. Currently we have:
"Placing after $r < n$ of the $n$ clones of strength $b$ finish:
$$\mathcal{L}(a)=\frac{b^ra}{(a+nb)(a+(n-1)b)(a+(n-2)b)\cdots(a+(n-r)b)}=\frac{B-1}{(B+n-1)(B+n-2)\cdots(B+n-r-1)}$$
where $B=1/b$ [also using the fact that $a+b=1$].
Making the change to $n-1$ clones of strength $b$ would give
$$\mathcal{L}(a)=\frac{b^ra}{(a+(n-1)b)(a+(n-2)b)(a+(n-3)b)\cdots(a+(n-r)b)}=\frac{B-1}{(B+n-2)(B+n-3)\cdots(B+n-r-2)}$$
which is a bit weird.
in
inst/very_simplified_likelihood.Rmd
we have the focal competitor andn
clones givingn+1
competitors. Maybe we should haven-1
clones, givingn
competitors. This would make things easier?