What is the difference between view and copy in NumPy?

Updated May 17, 2026

Short answer

A view shares memory; a copy duplicates data.

Deep explanation

Views reflect changes in original arrays because they share memory. Copies are independent and safe from side effects but cost more memory.

Real-world example

Avoiding unintended modifications in data pipelines.

Common mistakes

  • Assuming slicing always creates a copy.

Follow-up questions

  • How to check if array is view?
  • When should you use copy?

More NumPy interview questions

View all →