Estimating the Distribution of True Effects from Noisy Estimates

Estimates the distribution of true effects (or any latent parameter) from estimates that each have a known, possibly different, standard error. The main function, em_smooth(), implements the smoothed EM (EMS) algorithm of Silverman, Jones, Wilson and Nychka (1990) for a nonparametric mixing distribution on a grid. The smoothing uses one constant standard deviation, set from the typical standard error and the sample size.