Question

okay, I thought such code could not get wrong, but it obviously does:

somewhere:
    # p is a float value between 0 and 1
    m.limit=PidRange(p-1.0, p+1.0)

class PidRange(Range):
    def __init__(self, low, up):
        Range.__init__(low,up,...)
        pass
    # some methods definition for the PidRange sub-class

class Range(object):
    def __init__(self, p_min=None, p_max=None, ...):
        if (p_min > p_max):
             raise ValueError("Range can't be created: the low bound %f exceeds high bound %f."%(p_min,p_max))

I’m just trying to initialise a [min,max] range with some class hierarchy. But for some totally odd reason, p=0.888337 will
raise the following exception:

    File "src/__main__.py", line 155, in __find_data
        m.limit=PidRange(p-1.0, p+1.0)
    File "src/routing.py", line 32, in __init__
       Range.__init__(low, up, low!=None, up!=None)
    File "src/equation.py", line 30, in __init__
       raise ValueError("Range can't be created: the low bound %f exceeds high bound %f."%(p_min,p_max))
    ValueError: Range can't be created: the low bound 1.888337 exceeds high bound 1.000000.

Has anybody any clue about what’s happening ? I have to admit I’m far from mastering the Python language, but I fail to see any subtlety that could explain such an odd behaviour.

Answer 1

ah. Figured out.

• missing a self in __init__ call to the superclass
• 1.000 ‘extra value’ is a True from the … trans-typed to float

so superclass constructor invocation ‘breaks’ the call model and require explicit self reference, huh ?