Rank mapping of three-way multicomponent profiles

In this work, fast graphic procedures for assessment of number of analytes (chemical rank) in local regions of three-way unresolved multicomponent profiles are presented. Collapsing the three-way profile along one direction by matrix summation and evolving rank analysis on the resulting matrix represents a first step for the approach developed here. This procedure is efficient for detecting analytes with low net analytical signal compared to the noise level. The effect of collapsing is a reduction of random noise due to cancellations, while structure from analytes is enhanced because of the additive effect of many similar contributions. In the second step, slicing of the three-way data array into complete two-way arrays (matrices) is performed. Projections are then used across these matrices to generate a rank map. The strategy developed in this work provides a rapid approach to rank analysis and permits the analyst to take into account that the analytical information is normally not uniformly distributed in the three directions.