§ 1.6section 6 of 16·~9 min
Functions and handling bad input
Why functions
A function is a named, reusable block of code. It takes inputs (parameters) and gives back an output (return).
def gst(amount: float, rate: float = 0.18) -> float:
"""Return the GST on an amount."""
return round(amount * rate, 2)
gst(1000) # 180.0 (uses the default rate)
gst(1000, rate=0.05) # 50.0 (keyword argument: clear and safe)
Parts of a function:
def gst(...)gives it a name. Use verbs:add_expense,load_data.amount: floatis a parameter with a type hint. Python doesn't enforce hints, but they document your intent, and from Week 13 onwards LLMs read them.rate: float = 0.18is a default value, so callers may leave it out.-> floatsays what the function returns.- The
"""docstring"""is a one-line explanation that shows up when you hover over the name in VS Code. returnsends a value back. A function withoutreturnreturnsNone.
print vs return. Beginners often print inside functions. print shows text to a human, while return gives a value to the rest of your code. Prefer return and let the caller decide whether to print.
Scope. Variables created inside a function exist only inside it. Pass data in through parameters and get it out through return, rather than reaching for global variables.
*args and **kwargs (you'll see them everywhere)
def total(*amounts: float) -> float: # any number of positional arguments, as a tuple
return sum(amounts)
total(10, 20, 30) # 60
def make_expense(**fields): # any keyword arguments, as a dict
return fields
make_expense(amount=250, category="food") # {'amount': 250, 'category': 'food'}
Library code, including LangChain, often accepts **kwargs and passes them on to something else.
The mutable-default trap
def add_item(item, basket=[]): # BUG: the same list is shared by every call
basket.append(item)
return basket
add_item("chai") # ['chai']
add_item("vada") # ['chai', 'vada'] ← surprise!
def add_item(item, basket=None): # FIX
if basket is None:
basket = []
basket.append(item)
return basket