How to pass a function with some (but not all) arguments to another function?

Question

I have this function f(x, a, b):

def f(x, a, b):
    return a*x + b

and I have to pass this function to another one with 'a' and 'b' pre-set, e.g.:

def print2(f):
    x = np.linspace(0,1,10)
    print(f(x))

print2(f( , 1, 1))

I know that there are other ways to solve this particular problem, however this is just an example, not the real problem, and I'd like to know if there's anyway to do something like this "print2(f( , 1, 1))". In other words, is there anyway to do this without having to pass the arguments all the way through? Like,

def print3(f, a, b):
    x = np.linspace(0, 1, 10)
    print(f(x, a, b))
print3(f, a, b)

Answer 1

You can use functools.partial exactly for that!

def print2(f):
    x = np.linspace(0,1,10)
    print(f(x))
f2 = functools.partial(f, a=1, b=2)
print2(f2)

For the many possible advantages gained by using partial, see this.

In short:

• Partials allow introspection, you can know what you've set the values to.
• Partials are supported throughout the standard library in many areas and are recommended by modules such as asyncio.
• Partials can be pickled unlike lambda.
• Partials are fully fixed and don't need to use closure variables which might cause the fixation to change.
• Partials are faster to execute.

and many more...

Answer 2

You can just wrap it in another function:

g = lambda x: f(x, 1, 1)

print2(g)

Make the arguments of the outer function whatever data still needs to be supplied, and pass that data through.

---

I'm being told that this approach isn't ideal in Python. @Bharel's answer is likely more appropriate.

Answer 3

Both lambda and partial are fancy ways of creating a new function. You can also just write it yourself:

def f2(x):
    return f(x, 1, 1)

What makes lambda and partial special is that they are expressions and can be passed as arguments to higher-order functions directly.