Issue with Parsing ROOT:TFormula

[attikis]

Dear devs,

we are experiencing some trouble trying to implement a TFormula in correctionlib. Our understanding is that problem is related to the parsing of the "&&" logic operator in the TFormula.

cset = _core.CorrectionSet.from_file(fname)

RuntimeError: Failed to parse Formula expression at position 28:
((min(max(x,20.),140.)>20.0&&min(max(x,20.),140.)<=30.0))*0.8434999999999999
^
syntax error, unexpected '&', expecting , .

Can you please clarify if this indeed the case? If indeed the case do you think i'd be possible to rectify this?

In the meantime, is there a way to test a-priory a given TFormula or even parse it and then pass it to correctionlib?
Perhaps an example script or documentation?

Thank you in advance.

Best regards,
Alexandros

[nsmith-]

You can test formulas by making a small correction object in code:

from correctionlib import schemav2 as schema

c = schema.Correction(
    name="test",
    version=2,
    inputs=[schema.Variable(name="x", type="real")],
    output=schema.Variable(name="a scale", type="real"),
    data=schema.Formula(
        nodetype="formula",
        expression="23.*log(x)",
        parser="TFormula",
        variables=["x"],
    ),
).to_evaluator()

assert c.evaluate(1.0) == 0.0

Looking at your formula, it appears to be a binning. You should use a Binning node to implement this correction and not a formula.

[nsmith-]

Were you able to implement a binned formula?