Partial Function Application -- Advantages over normal function?

[Terry Reedy]

On 7/18/2011 8:24 AM, Paul Woolcock wrote:
> Partial function application (or "currying") is the act of taking a
> function with two or more parameters, and applying some of the arguments
> in order to make a new function.  The "hello world" example for this
> seems to be this:
>
> Let's say you have a function called `add`, that takes two parameters:
>
>> >> def add(left, right):
>      ...     return left + right
>
> Now let's say you want a function that always adds 2 to a number you
> give it.  You can use partial function application to do this:
>
>  >>> from functools import partial
>  >>> add2 = partial(add, right=2)
>
> Now, you have a new function, `add2`, that takes one parameter:
>
>  >>> add2(4)

Or you can directly write

def add2(x): return x + 2

or more generically

def makeadder(y)
     def _add(x): return x+y
add2 = makeadder(2)

functool.partial is essential a generic version of makeadder in that it 
also abstract the function/operator. It is useful when one has a 
function but perhaps not the source. It's limitation is that args are 
frozen left to right while the above example freezes the right operand.

-- 
Terry Jan Reedy