What is the need of defining the argument variable inside the function?

def load_housing_data(housing_path=HOUSING_PATH):
 

 csv_path = os.path.join(housing_path, "housing.csv")
 return pd.read_csv(csv_path)

Can't we simply ignore it like -

def load_housing_data():
 

  csv_path = os.path.join(HOUSING_PATH, "housing.csv")
  return pd.read_csv(csv_path)

I'm believe -

def load_housing_data():

is more right than -

def load_housing_data(housing_path=HOUSING_PATH):

CodePudding user response：

def load_housing_data(housing_path=HOUSING_PATH): argument housing_path have HOUSING_PATH as default value, that means, if you call function load_housing_data() and no passed that argument, by default taken HOUSING_PATH as value

CodePudding user response：

Both ways will work.

In the first one you are defaulting the argument to a value. This allows you to skip supplying one if the default value is sufficient, but still allows you to supply an alternate if it is not sufficient.

def load_housing_data(housing_path=HOUSING_PATH):
   csv_path = os.path.join(housing_path, "housing.csv")
   return pd.read_csv(csv_path)

#you can do this
data = load_housing_data()

#or this
data = load_housing_data(SOME_OTHER_PATH)

In the second you are statically supplying the value and removing your ability to supply an alternate one.

def load_housing_data():
    csv_path = os.path.join(HOUSING_PATH, "housing.csv")
    return pd.read_csv(csv_path)

#only option
data = load_housing_data()

However, if you are going to use the second version, then you might as well make csv_path global. You've made it so csv_path will never be different so, no reason to keep remaking it every time you call the def.

CSV_PATH = os.path.join(HOUSING_PATH, "housing.csv")

def load_housing_data():
    return pd.read_csv(CSV_PATH)

Once you do that, the def is essentially just an alias so, you don't need it.

CSV_PATH = os.path.join(HOUSING_PATH, "housing.csv")
data     = pd.read_csv(CSV_PATH)