Jason Willwerscheid

I think we should wrap the `mixcdf` results in a call to `drop` so that we return a vector rather than a 1 x k matrix.

The docs claim that it defaults to "uniform". In fact it defaults to "nullbiased".

doesn't handle zero sebetahats correctly

4

For example, `ash(betahat = c(1, 2, 3, 4), sebetahat = c(0, 0, 1, 1))` gives me posterior means of zero for the first two observations.

I think fixing this is as simple as changing `alpha = outer(-x, m$a,FUN="+")/s` to `alpha = outer(x, -m$b,FUN="+")/s` in `comp_postmean2` (since we send `-b` and `-a` to `e2truncgamma`) but I...

optimization method options should be better documented

4

even a basic description of what the options mean would be helpful; at present there is nothing