| False Discovery Rate using a Mixture Model Options |
| See Also |
| Maximum number of Iteration cycles | The maximum number of iterations of estimating the model parameters. If the model has not converged before this number of iterations the FDRMIXTURE procedure will exit with a warning. Increase this parameter to force more iterations. Changing the initial parameters may help decrease the number of iterations to convergence. Using the final parameter estimates as the initial settings on another run should increase the probability of convergence. i |
| Tolerance for convergence | When the all changes in current estimates of P, A, & B between iterations are smaller than this value, then the model fitting is deemed to have converged. If this value is too low, the model will take a long time to converge and the iterations will exceed the Maximum number of Iteration cycles. If this value is too large, the parameters of the model may still not be close to the optimum estimates when the model has deemed to have converged. |
| Parameter Estimates | The estimates of the P, A and B parameters in the mixture model |
| Monitoring of Fitting Iterations | The current estimates of P, A, & B and log-likelihood and changes in these on each iteration of the model fitting |
| Histogram of Probabilities & Model Fit | The plot of the fitted mixture against the histogram of probabilities |
| Density on the Logit scale | The fitted mixture against the kernel density estimate of the probabilities on a logit scale (this allows a more detailed comparison at small probability value). |
| Log Density on the Logit scale | As above, but gives even greater detail, by putting the density on a log scale (note that greater variation is expected around small density values on the log scale). |
| Inference (FDR, FRR, Power) | The plot of FDR, FRR and POWER against p |
| Inference (FDR, FRR, Power) on Log scale | Plots these statistics on log scales, with the X-axis restricted to probabilities < 0.5, with a background grid, to enable estimates to be read for specific probability values. |