Table 4 Comparison of the misclassification error between the raw datasets, the adjusted data using the default HarmonizR adjustment and two blocking approaches using a block size of 2 and 4 respectively. The tests were done on the datasets from Petralia et al. and Krug et al., using the KNN classifier provided by the Python package ’sklearn’. No prior sorting was performed

From: HarmonizR: blocking and singular feature data adjustment improve runtime efficiency and data preservation