Top 40 Python Interview Questions 2026 — With Crisp Answers

Basics (Every Interview)

• list vs tuple? — mutable vs immutable; tuples for fixed records/keys.
• What are decorators? — functions wrapping functions (@timer, @login_required).
• args vs kwargs? — *args positional tuple, **kwargs keyword dict.
• Shallow vs deep copy? — copy.copy shares nested refs; copy.deepcopy clones all.
• What is PEP 8? — style guide; mention flake8/black usage.
• is vs ==? — identity vs equality.

OOP + Intermediate

• self explained? — instance reference; first param of methods.
• Inheritance vs composition? — is-a vs has-a; prefer composition.
• What are generators? — lazy iterators via yield; memory-efficient loops.
• List comprehension vs map? — readability + speed for transforms.
• @staticmethod vs @classmethod? — no instance access vs class access.

Pandas + Data (Analyst Rounds)

• DataFrame vs Series? — 2D table vs 1D column.
• merge vs join vs concat? — SQL-style, index-based, stacking.
• How to handle missing values? — dropna/fillna by context; never silent mean-fill dates.
• groupby-apply pattern? — split-apply-combine; give a sales example.
• Vectorisation? — avoid row loops; use NumPy/Pandas ops.

Tricky + HR

• GIL? — one bytecode thread at a time; use multiprocessing for CPU work.
• Explain your best project in 2 minutes — problem → approach → metric → your role.
• Why data roles? — tie past + portfolio, not salary.

90-Day Prep Plan

Days 1–30: basics + OOP + 100 practice problems. Days 31–60: Pandas + 2 projects. Days 61–90: mocks (record yourself), revise GIL/decorators/generators weekly. Practice live: Python compiler. Full prep: ML interview prep.

10 More Rapid-Fire Questions (With Answers)

• Walrus operator? — := assigns inside expressions (if (n := len(x)) > 5).
• * unpacking? — a, *b = [1,2,3] → b=[2,3]; also f(*args).
• dict comprehension? — {k: v for k,v in pairs} — one-line dict builds.
• try/except/else/finally? — else runs if no exception; finally always (cleanup).
• Why virtualenv? — Isolated deps per project; avoids version clashes.
• __name__ == "__main__"? — True only when file runs directly — guards script code on import.
• Mutable default arg gotcha? — def f(x=[]) shares one list — use None + create inside.
• enumerate + zip? — Loop with index (enumerate) and parallel lists (zip).
• strip/split/join? — Clean, tokenise, rejoin strings — daily data cleaning trio.
• pip freeze? — Snapshots exact versions for requirements.txt reproducibility.

7-Day Revision Plan

• Day 1–2: definitions aloud (record yourself, 2-min each).
• Day 3–4: 1 project end-to-end narration.
• Day 5: full mock with a friend/mentor.
• Day 6: fix weak answers only.
• Day 7: rest + revise flashcards. Sleep beats cramming.

Annotated Mock Round (Analyst, 15 min)

• "Rate your Python." → Never "8/10". Say: "Strong in Pandas/SQL-automation; here's a 200-line EDA I shipped" + link. Why it works: proof over rating.
• Live: "Dedupe these customers, keep latest." → Think aloud: "Window ROW_NUMBER() in SQL, or drop_duplicates(subset, keep='last') in Pandas — I'd ask which source of truth." Scores process.
• "Your script is slow on 5M rows — fix?" → "Profile first; vectorise loops; chunk reads; push filters to SQL." Shows production thinking.
• "Explain groupby to a manager." → "Split rows into buckets, summarise each — like pivot tables that scale." Translation skill = promotion skill.

Frequently Asked Questions

• How many Python questions are asked in data interviews?
  8–15 per round: 40% basics/OOP, 40% Pandas, 20% project discussion.
• Is Python enough for data analyst interviews?
  Plus SQL + Power BI + 1 dashboard project — Python alone rarely suffices.
• Toughest Python topic?
  Decorators/generators + Pandas edge cases; practice out loud.
• Best way to revise before interview day?
  Your own project READMEs + 20 flashcards of definitions.

Read also: Learn Python · SQL Questions · Interview Guide