TTreeFormula behaves badly with <= operator and multiple conditions

[Timothy-Anders0n]

• Checked for duplicates

Describe the bug

Using pyRoot, preparing a selection string to use in TTree.GetEntries(const string * selection), it seems that a string of the type: _value1_ <= _variable_ <= _value2_"

should be working but insteads it returns an error: TTreeFormula bad expression "=_value2_"

when other operators work just fine (no error), e.g.:

• _value1_ < _variable_ < _value2_"
• _value1_ <= _variable_ < _value2_"
• etc.

I have tried with and without parentheses around the conditions. The only working solution (for now) to write: _value1_ <= _variable_ && _variable_ <= _value2_"

but this sensibly slows the code.

Expected behavior

Using _value1_ <= _variable_ <= _value2_" should not return an error when trying to limit a variable to a closed interval, in the same way it works when limiting a variable to an open interval.

To Reproduce

Steps to reproduce the behavior:

1. The code that can reproduce the issue is the following:

    import itertools
    import numpy as np
    import ROOT as rt
    paramList = { 'Variable_1' : np.linspace(0,1,4),
                          'Variable_2' : np.linspace(0,3,5) }

    boundaries = list(itertools.product(*[itertools.combinations(val,2) for val in paramList.values()]))
    for val in boundaries:
        string = 'dmID_signal == <whatever ID for final particle>'
        for bounds, key in enumerate(paramList.keys()):
            string += " && ({} <= {} <= {})".format(val[bounds][0], key, key, val[bounds][1])
        signal = sigData.GetEntries(string)

2. I cannot really join a .root file here, but it is just a matter of changing Variable_1 and Variable_2 of paramList for 2 variables of a TTree. Also, I start the selection string with a selection of dmID but that can be removed, you just need to modify the string so that it doesn't start with &&.

3. To run the code, write it in a python file or jupyter notebook and run it using Python 3.

Setup

1. JupyROOT 6.22/08
2. Linux Mint 19.3 Cinnamon
3. Installed Root using conda

Additional context

Using PyRoot