Random variates: Sometimes important. Easy to implement without starts for the probabilistic kernels (all except beta and gamma?) Can make a "conditional" function for this.
CDF: Again, probably easy for the probabilistic kernels. Will need numeric integration for the rest.
Quantile: Difficult! Will need some good ideas. Maybe the most useful of the bunch?
Should any of these be implemented?
Random variates: Sometimes important. Easy to implement without starts for the probabilistic kernels (all except beta and gamma?) Can make a "conditional" function for this.
CDF: Again, probably easy for the probabilistic kernels. Will need numeric integration for the rest.
Quantile: Difficult! Will need some good ideas. Maybe the most useful of the bunch?