choose_int.nl: Get intervention values
In neilbramley/acl_source: Functions for active causal learning

Computes the most useful intervention according to any desired function. Does not require likelihoods to be preloaded as it uses indi_li to compute these on the fly

1	choose_int.nl(DAGs, int_opts, prior, sar, pow, FUN, ...)

`DAGs`	an array where every slice is the next graph matrix in the hypothesis space
`int_opts`	the available interventions in a maatrix where each row is an intervention 1 is on, -1 is off, 0 is free
`prior`	the prior distribution over graphs
`sar`	spontaneous activation rate
`pow`	power of causal links
`FUN`	the function to use to evaluate the intervention options relative to the hypothesis space and prior

1
2
3

int_opts<-as.matrix(expand.grid(rep(list(0:2), 3)))
int_opts[int_opts==2]<--1
out<-choose_int.nl(DAGs_3, int_opts, prior('random',25), .1, .8, shannon_entropy)