How NumPy Helps You Land a Python Developer Job Faster

prices = [120.0, 80.0, 450.0, 999.0, 35.0]

discounted = []
for p in prices:
    if p >= 100:
        discounted.append(round(p * 0.9, 2))
    else:
        discounted.append(p)

expensive = [p for p in prices if p > 400]
print(discounted)
print(expensive)

import numpy as np

prices = np.array([120.0, 80.0, 450.0, 999.0, 35.0])

discounted = np.where(prices >= 100, np.round(prices * 0.9, 2), prices)
expensive = prices[prices > 400]

print(discounted)
print(expensive)
print("total:", prices.sum(), "average:", round(prices.mean(), 2))

import timeit
import numpy as np

rng = np.random.default_rng(0)
data = rng.random(1_000_000)
as_list = data.tolist()

def loop_version():
    return [x * 2 + 1 for x in as_list]

def numpy_version():
    return data * 2 + 1

assert np.allclose(loop_version(), numpy_version())
t_loop = min(timeit.repeat(loop_version, number=1, repeat=3))
t_np = min(timeit.repeat(numpy_version, number=1, repeat=3))
print(f"loop:  {t_loop:.4f}s")
print(f"numpy: {t_np:.4f}s")
print(f"speed-up on this machine: {t_loop / t_np:.1f}x")

import numpy as np

a64 = np.ones(1_000_000, dtype=np.float64)
a32 = np.ones(1_000_000, dtype=np.float32)
ints = np.arange(1_000_000, dtype=np.int32)

for name, arr in [("float64", a64), ("float32", a32), ("int32", ints)]:
    print(f"{name:8} {arr.nbytes / 1_000_000:.1f} MB")
# Smaller dtypes save memory but reduce precision; choose deliberately.

import numpy as np

# 3 students x 4 tests
scores = np.array([
    [70, 82, 91, 65],
    [88, 79, 85, 90],
    [60, 75, 70, 80],
])
print("shape:", scores.shape)
print("average per test (axis=0):", scores.mean(axis=0))
print("average per student (axis=1):", scores.mean(axis=1).round(1))

# Broadcasting: centre each test's scores around its mean
centred = scores - scores.mean(axis=0)
print(centred.round(1))

import numpy as np

# Min-max scale each column to 0-1 (a common interview task)
X = np.array([[1.0, 200.0], [2.0, 400.0], [3.0, 300.0]])
mins = X.min(axis=0)
ranges = X.max(axis=0) - mins
ranges[ranges == 0] = 1          # avoid division by zero
scaled = (X - mins) / ranges     # broadcasting (3,2) with (2,)
print(scaled)

Project: Fast analytics module

- Python package with functions for moving averages, z-scores,
  min-max scaling and outlier flags, all vectorised with NumPy
- Loop versions kept for comparison
- Benchmarks with timeit on several data sizes
- pytest tests checking results with np.allclose
- README: what each function does, measured speed-ups, limits

import numpy as np

def moving_average(values, window: int) -> np.ndarray:
    arr = np.asarray(values, dtype=float)
    if window < 1 or window > arr.size:
        raise ValueError("window must be between 1 and len(values)")
    kernel = np.ones(window) / window
    return np.convolve(arr, kernel, mode="valid")

print(moving_average([1, 2, 3, 4, 5, 6], 3))   # [2. 3. 4. 5.]

Week 1  Arrays, dtypes, indexing, slicing, boolean masks
Week 2  Axes, broadcasting, reshape, np.where, aggregations
Week 3  Benchmarking, tests, packaging the project with a README