Drag & drop to use
Drag & drop this node right into the Workflow Editor of KNIME Analytics Platform (4.x or higher).Learn more
For each chosen column (parameter) a reference population is used to estimate n different bins. The bins are based on the quantiles of the data. Then all data is grouped by the aggregation label (e.g. well) and the data of each group is applied to the bins. The percentage of data falling into a certain bin is calculated and used to estimate a z-score. The result contains a z-score, a percentage and an absolute count for each bin of each aggregation group and each parameter. It should help to detect minor distribution changes which would not be caught with a mean or median per aggregation group of the object data.