wR

wR is a shorthand used in various disciplines to denote a weighted version of a residual, regression, or error metric. Its exact meaning depends on the context and field of study, so careful interpretation is required.

In statistics and data analysis, wR commonly denotes weighted regression or weighted residuals. In weighted least squares (WLS), observations are assigned weights that reflect their variances or importance, leading to a weighted regression model and associated residuals. Weighted residuals and related diagnostics are used to assess model fit, particularly when data heterogeneity or varying measurement precision is present. Software implementations may report wR as part of fit quality or diagnostic outputs.

In X-ray crystallography and related structural biology fields, wR refers to the weighted R-factor, a goodness-of-fit statistic used during refinement. The weights modulate the contribution of each measured reflection to the overall agreement between observed and calculated structure-factor amplitudes. Lower wR values indicate a better fit, but the metric is sensitive to the weighting scheme and data quality, and it is not directly comparable across different experiments without consistent weighting and data processing.

Other uses of wR may appear in different domains as an abbreviation for weighted residuals, weighted regression, or related weighted metrics. Because wR has field-specific meanings, it is advisable to consult domain-specific references to determine the precise definition in a given context.