CODEWITHSUNDEEP
pythontypingabstract-base-class
phytebunk
phytebunk

Reputation: 13

Python: Implement base class for different types

I wonder if there is a way to implement subclasses of a base class for different types. Each subclass should have individual input and output types while providing same behaviour as the base class.

Background: I want to process voltage and temperature samples. 100 voltage samples form a VoltageDataset. 100 temperature samples form a TemperatureDataset. Multiple VoltageDatasets form a VoltageDataCluster. Same for temperature. The processing of Datasets depends on their physical quantity. To ensure that voltage related processing can't be applied to temperature samples I'd like to add type hints.

So I'd would be nice if there is a way to define that VoltageDataClustes method append_dataset allows VoltageDataset as input type only. Same for temperature.

Is there a way to implement this behaviour without copy&pasting?

# base class
class DataCluster:
    def __init__(self, name):
        self.name = name
        self.datasets = list()

    def append_dataset(self, dataset: Dataset) -> None:
        self.datasets.append(dataset)

# subclass that should allow VoltageDataset input only.
class VoltageDataCluster(DataCluster):
    pass

# subclass that should allow TemperatureDataset input only.
class TemperatureDataCluster(DataCluster):
    pass

Thanks! Niklas

Upvotes: 1

Views: 505

Answers (2)

hnfl
hnfl

Reputation: 289

You could use pydantic generic models.

from typing import Generic, TypeVar, List
from pydantic.generics import GenericModel

DataT = TypeVar('DataT')

class DataCluster(GenericModel, Generic[DataT]):
    name: str
    datasets: List[DataT] = []

    def append_dataset(self, dataset: DataT) -> None:
        self.datasets.append(dataset)

voltage_cluster = DataCluster[VoltageDataset](name="name")
voltage_cluster.append_dataset(some_voltage_dataset)

Upvotes: 2

Marko Delic
Marko Delic

Reputation: 96

When you inherit a class it automatically inherits the functionality of the class so there is no need to copy and paste. I'll illustrate this with an example.

# DataCluster.py
class DataCluster:
  def __init__(self, name):
    self.name = name

  def printHello(self):
    print("Hello")
  # This will work in sub classes that have a "data" attribute
  def printData(self):
    print(self.data)


# VoltageDataCluster.py
from superclasses.DataCluster import DataCluster
class VoltageDataCluster(DataCluster):

  def __init__(self, differentInput):
    self.differentInput = differentInput
    self.data = "someotherdata"


# mainclass.py
from superclasses.DataCluster import DataCluster
from superclasses.VoltageDataCluster import VoltageDataCluster

try:
  dc = DataCluster("mark")
  dc.printHello();

  # The input for this class is not name
  vdc = VoltageDataCluster("Some Other Input")

  # These methods are only defined in DataCluster
  vdc.printHello()
  vdc.printData()

As you can see, even though we only defined the "printHello" method in the super class, the other class inherited this method while using different inputs. So no copy and pasting required. Here is a runnable example (I added comments to tell you where to find each file used).

EDIT: Added a data attribute so its more relevant to your example.

Upvotes: 0

Related Questions