Search code examples
pythonoverriding

How can I avoid code repetition when overriding many properties in a child class in the same way?


I have a class in which I have properties that are returning arrays. For simplicity, let's consider them constant:

import numpy as np

class MyClass:
    def __init__(self):
        self._time = np.array([0, 1, 2, 3])
        self._a = np.array([0, 1, 2, 3])
        self._b = np.array([4, 5, 6, 7])

    @property
    def a(self):
        return self._a

    @property
    def b(self):
        return self._b

Now, I have another class which is inheriting MyClass and it is interpolating the data, for example:

class Interpolator(MyClass):
    def __init__(self, vector):
        super().__init__()
        self._vector = np.array(vector)

    @property
    def a(self):
        return np.interp(self._vector, self._time, self._a)

    @property
    def b(self):
        return np.interp(self._vector, self._time, self._b)

Now, the issue is that I have 2 classes like MyClass and each one of them consists of ~30 properties.

Is there a way to override all properties without doing it one by one? I was having a look also at this solution but I am not sure if/how I can apply it to my problem.


Solution

  • Refactor your superclass to "proxy"/"trampoline" those properties via a function you can override in a subclass:

    import numpy as np
    
    
    class MyClass:
        def __init__(self):
            self._time = np.array([0, 1, 2, 3])
            self._a = np.array([0, 1, 2, 3])
            self._b = np.array([4, 5, 6, 7])
    
        def _get_property(self, v):
            return v
    
        @property
        def a(self):
            return self._get_property(self._a)
    
        @property
        def b(self):
            return self._get_property(self._b)
    
    
    class Interpolator(MyClass):
        def __init__(self, vector):
            super().__init__()
            self._vector = np.array(vector)
    
        def _get_property(self, v):
            return np.interp(self._vector, self._time, v)