Python reduce() Deep Dive — Accumulate One Result with functools

What you will learn

• Why reduce is not a built-in in Python 3 — import functools
• Syntax varname = functools.reduce(FunctionName, iterable)
• The notes' four internal steps (first two values, then the next, until one result)
• Sum, max, min, and word concatenation
• def operation(m, p) vs lambda m, p: m + p

Functions showed a first reduce. This lesson follows reduce().txt and ReduceFun(s)Ex1–5.

What is reduce()?

reduce() combines all elements of an iterable into one final value. It is not a built-in function in Python 3, so it must be imported from the functools module.

Import syntax:

import functools
result = functools.reduce(FunctionName, iterable)

Syntax breakdown

The callback always receives two arguments: m (accumulator) and p (next value).

Unlike filter() and map(), the callback takes two parameters.

Four internal steps

The teaching notes describe reduce() as repeatedly combining two values until only one result remains.

Process:

1. Take the first two values from the iterable (m then p) and pass them to FunctionName. Store that result internally.
2. Take the next unused value as the new p. The previous result becomes m. Call FunctionName again.
3. Repeat until every element has been used.
4. The last internal result is returned — one value.

Example: [10, 20, 30, 40] with add:

• 10 + 20 → 30
• 30 + 30 → 60
• 60 + 40 → 100

Sum with lambda or def

reduce() is commonly used to calculate totals.

import functools

lst = [10.0, 20.0, 30.0, 40.0, 50.0]
res = functools.reduce(lambda m, p: m + p, lst)
print("Sum({})={}".format(lst, res))  # 150.0

Lambda: lambda m, p: m + p · Output: 150.0

Lambda vs a normal function

Both implementations produce the same result.

Named function (ReduceFunsEx2):

import functools as fc

def operation(m, p):
    return m + p

res = fc.reduce(operation, lst)

Both callbacks always accept (m, p). That is different from filter / single-iterable map. There is no reduce(...) in the builtin namespace.

Finding maximum

Instead of using Python's max(), reduce() keeps the larger value at every step.

Logic: lambda m, p: m if m > p else p · Final result: 56.0

values = [10.0, 2.0, 12.0, 3.0, 45.0, 4.0, 56.0, 13.0, -3.0, 6.0]
val = functools.reduce(lambda m, p: m if m > p else p, values)
print("max=", val)  # 56.0

Each step keeps the winner and compares it to the next element.

Finding minimum

The same idea works for the smallest value by reversing the comparison.

Logic: lambda m, p: m if m < p else p · Final result: -3.0

val = functools.reduce(lambda m, p: m if m < p else p, values)
print("min=", val)  # -3.0

Concatenate words

reduce() can also combine strings into a single sentence.

Lambda: lambda m, p: m + " " + p

words = ["Python", "filter", "map", "reduce"]
line = functools.reduce(lambda m, p: m + " " + p, words)
# Python filter map reduce

Or a def:

def concatop(m, p):
    return m + " " + p

line = functools.reduce(concatop, words)

Classroom code splits comma-separated input(). The playground uses a list literal.

Common mistakes

Remember:

• Import functools — there is no builtin reduce(...)
• Callback needs two parameters (m, p), not lambda x: x + 1
• reduce() returns one value, not a list
• Add " " when joining words — m + p glues tokens together

Complete reduce() pipeline

This summarizes how accumulation works from start to finish.

Interview cheat sheet

Topic	Answer
Module	functools
Syntax	functools.reduce(fn, iterable)
Parameters	Two → m, p
First Call	Uses the first two elements
Return Value	One accumulated result
Applications	Sum, Max, Min, String Concatenation
Alias	import functools as fc

Practice in the playground

1. Reduce a product: lambda m, p: m * p on [1, 2, 3, 4].
2. Concatenate with a hyphen instead of a space.
3. Drop as fc and call functools.reduce only.

Next: Filter → Map → Reduce pipeline.