Currently the numerical test for the mean estimation of the Beta distribution may be small enough that a small sample size would fail to estimate it. This does not always happen but it's hard to know when it's going to fail.
Reduce the tolerance so a reasonable sample size can estimate it well.
Currently the numerical test for the mean estimation of the Beta distribution may be small enough that a small sample size would fail to estimate it. This does not always happen but it's hard to know when it's going to fail.
Reduce the tolerance so a reasonable sample size can estimate it well.