The script receives 9 arguments:

• evaluation: User evaluation to inspect

• class-name: Name of the positive class, or the class to inspect

• cost-FP: False positive cost

• cost-FN: False negative cost

• cost-TP: True positive cost

• cost-TN: True negative cost

• benefit-TP: True positive benefit

• perc-retention: Percentage (0-1) of retention in TPs

• dataset-name: Name of the resources (Source inline and Dataset)

And returns as output a dataset-id which header is: "Threshold,TP,FP,TN,FN,Cost_Fx,Cost_Benefits"