search

stephenslab / ebnm

R package to fit Empirical Bayes Normal Means model.
https://stephenslab.github.io/ebnm/
12 stars 9 forks source link

correct lfsr for point_normal, point_laplace, etc #45

Closed willwerscheid closed 3 years ago

willwerscheid commented 3 years ago

these currently give the "false sign" rate with respect to the mode rather than zero, so a kind of over-under moreso than a false sign rate -- we should either change the terminology when mode != 0 or change the logic; as is, the behavior is different from normal_scale_mixture and ashr

willwerscheid commented 3 years ago

I'll take the easy way out and not return the LFSR (with a warning) when the mode is not zero.