How to implement a list of objects as an instance attribute in Python?

Question

I am new to the OOP with Python. Until now I have worked with OOP languages as C++, C#, and Java. Since I have to translate a C# module to Python, I encountered conceptional differences.

I have this C# class relationship:

 internal class ChirpMeasDataClass
  {
     public ChirpMeasDataClass(List<AntennaLevelDataClass> AntennaLevelData, List<DiffphaseDataClass> DiffphaseData)
     {
        this.AntennaLevelData = AntennaLevelData;
        this.DiffphaseData = DiffphaseData;
     }
     internal List<AntennaLevelDataClass> AntennaLevelData = new List<AntennaLevelDataClass>();
     internal List<DiffphaseDataClass> DiffphaseData = new List<DiffphaseDataClass>();
  }
  internal class AntennaLevelDataClass
  {
     internal double dLevel =  -10000;          // [dB]
     internal double dVarianzLevel = -10000;  // [dB^2]
  }
  internal class DiffphaseDataClass
  {
     internal double dDiffPhase = -10000;         // [°] 
     internal double dVarianzDiffPhase = -10000; // [°^2] 
  }

What intuitively commes to mind in python is to have something like this:

class AntennaLevelDataClass:
    def __init__(self):
        self.dLevel = -10000  # [dB]
        self.dVarianzLevel = -10000  # [dB ^ 2]

class DiffPhaseDataClass:
    def __init__(self):
        self.dDiffPhase = -10000 # [°]
        self.dVarianzDiffPhase = -10000 # [° ^ 2]

class ChirpMeasDataClass:
    def __init__(self,AntennaLevelDataClass,DiffPhaseDataClass):
       self.antennaLevelDataList=AntennaLevelDataClass
       self.diffPhaseDataList=DiffPhaseDataClass

or something like this :

   class ChirpMeasDataClass:
        def __init__(self,antennaLevelDataClass,diffPhaseDataClass):
           self.antennaLevelDataList=[]
           self.diffPhaseDataList=[]

What is the common way to have list of Objects as an attribute in a class (instance attribute in Python)?

Python isn't typed (footnote, asterisk, it has optional type annotations now…), so the only thing you do is take an argument in the constructor and assign it to a property. That's it. That argument is supposed to be a list containing only instances of a specific type, but all that is rather irrelevant and doesn't need to be declared in any specific way. So, def __init__(self, a, b): self.a = a; self.b = b is really all you need. — deceze
– deceze ♦, Commented Sep 4, 2018 at 9:38

deceze · Accepted Answer · 2018-09-04 09:46:59Z

1

The only thing that C# constructor does is assign two arguments, which are lists of objects, to two instance properties. The rest is largely type annotations, which are entirely optional in Python. So, the basic thing you want is just:

class ChirpMeasData:
    def __init__(self, antenna_level_data, diff_phase_data):
       self.antenna_level_data = antenna_level_data
       self.diff_phase_data = diff_phase_data

You could add type annotations:

from typing import List

class ChirpMeasData:
    def __init__(self, antenna_level_data: List[AntennaLevelData],
                 diff_phase_data: List[DiffphaseData]):
       self.antenna_level_data = antenna_level_data
       self.diff_phase_data = diff_phase_data

(Note that I'm foregoing the redundant use of "Class" in the class names here.)

answered Sep 4, 2018 at 9:46

deceze♦

525k89 gold badges806 silver badges954 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

RemcoGerlich · Accepted Answer · 2018-09-04 09:41:50Z

0

If you do

self.antennaLevelDataList=AntennaLevelDataClass

Then self.antennaLevelDataList is an instance variable, and its value is that class definition. That doesn't seem very useful.

Moreover, you're passing two class definitions into the constructor.

Using lists (of which the contents happen to be instance of those classes) is much cleaner, and C++ is also using Lists so it's mostly a direct equivalent. The only difference is that in C++ you have to declare what type the contents of the list are and in Python you can't.

In the constructor, it's probably useful to copy the contents of the lists that are passed in, something like:

class ChirpMeasDataClass:
    def __init__(self, antennaLevelData, diffPhaseData):
        self.antennaLevelDataList = anntennaLevelData[:]
        self.diffPhaseDataList = diffPhaseData[:]

([:] indicates a slice of the entire list, in effect a shallow copy)

The point is that otherwise the instance attributes refer to the same list object as the ones passed in, and that might lead to surprises if one is copied; nothing is every copied implicitly in Python.

answered Sep 4, 2018 at 9:41

RemcoGerlich

31.4k6 gold badges66 silver badges83 bronze badges

3 Comments

deceze Over a year ago

"and its value is that class definition" – OP is shadowing the name AntennaLevelDataClass with a parameter there…

deceze Over a year ago

"and in Python you can't" – …and in Python it's optional [and not enforced by the runtime].

deceze Over a year ago

"nothing is every copied implicitly in Python" – Right, which is why implicitly making a copy of the list in the constructor is arguably the surprising behaviour.

Collectives™ on Stack Overflow

How to implement a list of objects as an instance attribute in Python?

2 Answers 2

Comments

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related