We need to know if the new diagrams are valid, i.e., without any numerical artefacts. Therefore, we can check:
(i) If a new diagram (consisting of reconstructed time series) havs the peaks just existing in the original series (i.e. no important artificial peaks are generated numerically).
(ii) A sum of all new separated diagrams should be equal to the original diagram. The di erence between these two diagrams can exhibit negligible WASEM numerical interferences.