SensitivityNormDatasets: Computes sensitive to normalization for dataset sources
In sevvandi/outselect: Algorithm selection for unsupervised outlier detection

Computes sensitive to normalization for dataset sources

1	SensitivityNormDatasets(rocpr = 1, xi = 0.05)

`rocpr`	If `rocpr=1` then area under ROC curve is used as the performance measure. If `rocpr=2` area under PR curve is used.
`xi`	The xi-sensitivity to normalization parameter xi. Defaults to `0.05`

A list containing the following:

KruskalWallis: The output of the KruskalWallis from stats::kruskal.test.
max_val: The dataset source, which is most sensitive to normalization.
min_val: The dataset source, which is least sensitive to normalization.
xi_sensitive_num: For how many outlier methods each dataset is sensitive to normalization.