A set of Free-R reflections may be selected at random for a given set of observations. The proportion of the reflections assigned to the Free-R set defaults to 5%, but can be altered. The Free-R set are selected in the highest symmetry point group consistent with the current point group, in case the spacegroup has been wrongly determined. The Free-R set is described by a data object containing flags which identify which reflections belong to the Free-R set.
To generate a Free-R set for a given set of observations, all that is required is the reflection data object for those observations.
Sometimes it is necessary to extend an existing set of Free-R flags to cover additional reflections, for example if a new set of observations is available at a higher resolution. In this case the same Free-R set should be retained for the existing reflections to avoid biasing the Free-R factor. In this case the 'complete an existing set of FreeR flags' checkbox should be checked, and the existing Free-R set object selected.
There is one option which allows the proportion of data in the free-R set to be selected. By default this is set to 5% of the data (a value of 0.05), however by checking the box a different proportion may be selected. For very small structures, a larger value may be required to provide sufficient reflections in the Free-R set in any given resolution shell.