Session 24 · Phase 3: Python

Python Practice Marathon — 30 Coding Problems

Consolidate everything from Sessions 16–23. Thirty problems spanning Python basics, NumPy, Pandas, and visualization — try each before revealing the solution.

⏱ ~3 hrs 🧠 Practice 🎯 Highest priority

How to Use This Marathon

  1. Type it, don't read it. Run every snippet — muscle memory beats recognition.
  2. Predict before you run. Write down the expected output first.
  3. Narrate out loud. This is exactly what a live-coding round feels like.

Part A — Python Basics (Problems 1–5)

1. Print each number in [10, 20, 30, 40] using a for loop. 
for n in [10, 20, 30, 40]:
    print(n)
2. Write a function that returns "High"/"Medium"/"Low" for a balance. 
def categorize(balance):
    if balance >= 100000:
        return "High"
    elif balance >= 30000:
        return "Medium"
    return "Low"
3. Use a list comprehension to get the squares of 0–9. 
squares = [x**2 for x in range(10)]
4. Create a dictionary and print only the keys whose values exceed 50000. 
balances = {"Aarav": 25000, "Priya": 120000, "Rahul": 8000}
for k, v in balances.items():
    if v > 50000:
        print(k)
5. Remove duplicates from a list using a set. 
cities = ["Mumbai", "Delhi", "Mumbai", "Chennai"]
unique = list(set(cities))

Part B — NumPy (Problems 6–10)

6. Create a NumPy array [10, 20, 30, 40, 50] and compute its mean and max. 
import numpy as np
arr = np.array([10, 20, 30, 40, 50])
print(arr.mean())   # 30.0
print(arr.max())    # 50
7. Multiply every element of that array by 2 (vectorized). 
arr * 2   # [20 40 60 80 100]
8. Reshape np.arange(6) into a 2×3 array. 
np.arange(6).reshape(2, 3)
9. Select the first three elements of the array via slicing. 
arr[0:3]   # [10 20 30]
10. Add 100 to every element (broadcasting). 
arr + 100   # [110 120 130 140 150]

Part C — Pandas: Creating & Selecting (Problems 11–15)

11. Create a DataFrame with columns product, price, units. 
import pandas as pd
df = pd.DataFrame({
    "product": ["Pen", "Notebook", "Eraser"],
    "price":   [25, 120, 10],
    "units":   [10, 5, 20]
})
12. Read a CSV file into a DataFrame. 
df = pd.read_csv("sales.csv")
13. Print the shape, columns, and first 5 rows. 
print(df.shape)
print(df.columns)
print(df.head())
14. Select the price column using both loc and iloc. 
df.loc[:, "price"]   # label-based
df.iloc[:, 1]        # position-based
15. Select the first row using both loc and iloc. 
df.loc[0]    # label
df.iloc[0]   # position

Part D — Pandas: Filtering & Cleaning (Problems 16–20)

16. Count missing values per column. 
df.isnull().sum()
17. Drop rows with any missing value. 
df.dropna()
18. Fill missing values in the price column with the column mean. 
df["price"] = df["price"].fillna(df["price"].mean())
19. Remove duplicate rows. 
df = df.drop_duplicates()
20. Filter rows where balance is greater than 50000. 
df[df["balance"] > 50000]

Part E — Pandas: groupby & Merge (Problems 21–25)

21. Total sales per region using groupby. 
df.groupby("region")["sales"].sum()
22. Sum and mean sales per region using agg. 
df.groupby("region")["sales"].agg(["sum", "mean"])
23. Merge two DataFrames on a shared key column (left join). 
pd.merge(orders, customers, on="customer_id", how="left")
24. Concatenate two DataFrames vertically. 
pd.concat([df1, df2])
25. Create a pivot table of sales by region (rows) × product (columns). 
df.pivot_table(index="region", columns="product",
               values="sales", aggfunc="sum")

Part F — Visualization & Mini-EDA (Problems 26–30)

26. Create a line plot of sales over time. 
import matplotlib.pyplot as plt
plt.plot(months, sales)
plt.show()
27. Create a bar chart of sales by region. 
plt.bar(regions, sales_by_region)
plt.show()
28. Create a histogram of the sales column. 
plt.hist(df["sales"], bins=10)
plt.show()
29. Create a Seaborn countplot of the region column. 
import seaborn as sns
sns.countplot(x="region", data=df)
plt.show()
30. Boxplot of sales by region, plus describe(). 
sns.boxplot(x="region", y="sales", data=df)
plt.show()
df.describe()
📋 Stable content — Reviewed: August 2026

Interview Questions — Python Live-Coding Classics

The Python questions interviewers ask again and again. Self-test before revealing.

IQ1. What are the core Python skills a data analyst needs?

Model answer: "Python basics (data structures, loops, functions), NumPy for arrays, Pandas for tabular data — selecting, cleaning, grouping — and Matplotlib/Seaborn for visualization."

IQ2. How do you approach a Python live-coding question?

Model answer: "I restate the task, outline my approach in words, write the code, then walk through the expected output. I'd rather narrate a clear plan than rush to code silently."

IQ3. Which library would you use to clean and aggregate tabular data?

Model answer: "Pandas — I'd use read_csv to load it, dropna/fillna for missing values, drop_duplicates for duplicates, and groupby to aggregate."

IQ4. How do you decide between a loop and a vectorized operation?

Model answer: "For numerical work on arrays, I use vectorized NumPy/Pandas operations — they're faster and cleaner. I use a loop only when the logic can't be expressed as a vectorized operation."

IQ5. What's the difference between dropna and fillna?

Model answer: "dropna removes rows/columns with missing values; fillna replaces them with a value. I choose based on how much data I'd lose and why it's missing."

IQ6. How do you group data and aggregate in Pandas?

Model answer: "With df.groupby('col')['metric'].sum() — or .agg(['sum','mean']) for multiple stats. It's the Pandas equivalent of SQL GROUP BY."

IQ7. How do you visualize results after analysis?

Model answer: "With Matplotlib for basic plots — line, bar, histogram, scatter — and Seaborn for statistical plots like countplot, boxplot, and heatmap."

IQ8. Walk me through a simple end-to-end analysis in Python.

Model answer: "Load with read_csv → inspect with shape/head/info → clean missing/duplicates → summarize with describe and groupby → visualize distributions and relationships → write insights."

💡
Pattern recognition: 80% of Python data interviews test the same eight skills — data structures, loops, functions, NumPy arrays, Pandas selecting, cleaning, grouping, and visualization. Master these and you've covered nearly everything.

Key Takeaways

1

Type every snippet — Python is learned by running, not reading.

2

Vectorize with NumPy/Pandas instead of looping where possible.

3

loc = labels, iloc = positions; groupby = aggregate; merge = join.

4

Narrate your approach out loud in a live-coding round.

5

You've completed the full Python stack — ready for any Python round.

Objective Questions — Test Your Understanding

Q1. Which expression builds a list of squares of 0–9?

Q2. Which method removes rows with missing values in Pandas?

Q3. Which method groups data by a column and aggregates?

Q4. Which Matplotlib function creates a line plot?

Q5. Which Pandas function combines DataFrames on a shared key column?