NumPy is the undisputed backbone of Python's data stack. While most developers know how to initialize arrays, compute sums, and reshape matrices, NumPy hides a wealth of advanced, high-performance features under the hood.
If you want to write faster, cleaner, and more expressive vectorised code, here are 5 powerful NumPy tricks you should add to your toolkit today.
1. Zero-Copy Sliding Windows with sliding_window_view
When working with time-series data, audio processing, or rolling statistics, you often need to create sliding windows over an array. Historically, this required messy index manipulation or memory-heavy loops.
NumPy provides a built-in view utility that creates sliding windows without copying any memory:
import numpy as np data = np.array([1, 2, 3, 4, 5, 6]) # Create rolling windows of size 3 windows = np.lib.stride_tricks.sliding_window_view(data, window_shape=3) print(windows) # Output: # [[1 2 3] # [2 3 4] # [3 4 5] # [4 5 6]]
