Lists & Tuples: Memory Layout, Slicing & Performance

1. How CPython Implements Lists (Dynamic Over-Allocation)

In CPython, a list is not a linked list; it is a contiguous dynamic array of pointersto other PyObject instances on the heap. When appending elements, resizing the array on every insertion would result in terrible O(N^2) performance.

Instead, CPython uses an over-allocation formula (approx. new_allocated = (size + (size >> 3) + 6) & ~3). This ensures that list.append() achieves an amortized O(1) constant time complexity!

import sys

# Observe dynamic over-allocation in real time
items = []
print(f"Empty list initial bytes: {sys.getsizeof(items)}")

for i in range(12):
    items.append(i)
    # Notice bytes jump in discrete chunks (over-allocation) rather than per element!
    print(f"Length: {len(items):2d} | Allocated Memory: {sys.getsizeof(items):3d} bytes")

2. List vs Tuple: Architectural Differences

3. Advanced Slicing Syntax: [start:stop:step]

Slicing allows you to extract sub-sequences cleanly. Slices are half-open intervalsincluding start up to but excluding stop:

data = [0, 10, 20, 30, 40, 50, 60, 70, 80, 90]

print(data[2:6])    # [20, 30, 40, 50] (Elements from index 2 to 5)
print(data[::2])    # [0, 20, 40, 60, 80] (Every 2nd element)
print(data[::-1])   # [90, 80, 70, 60, 50, 40, 30, 20, 10, 0] (Reverse list)
print(data[-3:])    # [70, 80, 90] (Last 3 elements)

4. Self-Documenting Data Records with typing.NamedTuple

Instead of accessing tuple elements using error-prone integer indices (record[0], record[1]), use typing.NamedTuple to create memory-efficient, self-documenting data classes:

from typing import NamedTuple

class GeoCoordinate(NamedTuple):
    latitude: float
    longitude: float
    elevation_m: float = 0.0

coord = GeoCoordinate(latitude=37.7749, longitude=-122.4194)

# Access by readable attribute name OR by index!
print("Latitude:", coord.latitude)
print("Index 0:", coord[0])

# Unpacking works seamlessly
lat, lon, elev = coord
print(f"Unpacked location: {lat}, {lon}")