Home

Last updated: 2016-10-25

Code version: 3358c823d6eba747cc51deae81eb95e66147cee0

Simple illustrations

• Figures illustrating FDR
• Illustatrate dilution effect of low-precision observations
• Adaptive Shrinkage

• Adaptive Shrinkage: shrunken interval estimates

• Comparing models: EE vs ET models
• Checking how qvalue behaves in simulation suggested by referee
• Comparison of methods, and example of Credible Interval behaviour when UA does not hold

• Using ash to fit the NPMLE; comparison with REBayes

DSC Shrink (the main simulation study in the paper)

• Plot Densities of simulations
• Plot estimates of pi0
• Plot local false sign rates and lfdr vs truth

• Plot CDFs

• Investigate coverage of hu method without penalty
• Run mixfdr on low noise bimodal situation

• Make coverage tables

Optimization method: EM vs Interior Point

• Compare loglik achieved by EM vs Interior Point
• Compare speed of EM vs Interior Point

DSC Znull (assess behaviour of ash LR under null)

• Summarize results of null simulations

Check how estimating the mode performs

• Basic check of non-zero model option