The phase comparison task allows phase estimates from different sources to be compared.
A set of reflection data and two sets of phase probabilities must be provided. The phases will be compared. The reflection data are used to calculate map correlations for maps calculated using the two phase sets.
The mean phase error between the datasets is reported. A table of addition statistics, along with graphs against resolution, are provided.
Weighted mean phase error includes a figure-of-merit to weight the phase comparison. This avoids the phase error being inflated by reflections for which no phase error is present, and also provides a suggestion as to the meaningfulness of the figure of merit. Weighting may be done using the figure of merit from either set of phase probabilities.
Map correlations further introduce the reflection magnitude into the weighting calculation. This avoids the phase error being inflated by reflections for which there is little or no intensity. F map correlation tends to be dominated by phase errors at low resolution, whereas E-map correlation is dominated by phases at high resolution.