Three command line tools:

• mvaExtractor constructs trainer description template from existing calibration object (only structure, not complete) First argument is an MVA file, the second one is the output XML file.
• mvaTreeComputer runs over many root trees, computes discriminator for each entry using MVA calibration file, stores result together with copy of original branches in new root file. MVA calibration and output root file 1st and 2nd argument
• mvaTreeTrainer reads in XML trainer description file, at least one ROOT tree, runs MVA trainer on the entries, writes out MVA calibration file with the results