Skip to main content
. 2018 Sep 3;74(Pt 9):877–894. doi: 10.1107/S2059798318009191

Figure 7.

Figure 7

Time-dependent ensemble refinement and integration results. All lattices from two groups of data on the front detector (collected at distances of 130 and 105 mm) were separated chronologically into runs and then subdivided chronologically into batches of 3000–4500 images. The detector position and crystal models were refined for each batch separately and the data were then integrated. (a) Refined detector distances for each batch plotted versus run number. Black line: starting detector distance before refinement. The change in detector distance needed to effect a one-pixel radial shift for a reflection at 2.2 Å is shown as a vertical gray bar. (b) Percent change in signal strength 〈Ic(I)〉 [σc(I) refers to the counting error, or the uncorrected error from integration summation] after per-run ensemble refinement. The mean resolution-binned signal after refinement for each batch was divided by the mean resolution-binned signal before refinement for that batch. Each line represents one batch. Percent change in signal is reported as a function of resolution. The numbered circles in (a) group together batches that have similar refined distances. These group numbers are also used to label the batches in (b).