Ground 0 → Expert Point 100

Comprehensive assignment pack for NumPy Special

This assignment pack is layered. It starts with simple array creation and ends with expert-style numeric thinking, including broadcasting, aggregation, matrix work, timing comparison, and a mini project that feels close to real AI-ML preparation.

How to use this pack

  • Students should attempt levels in order.
  • Each level increases conceptual depth, not only code length.
  • Encourage written explanation with every solution.
  • Students should show outputs, not only paste code.

Submission rule

  • One file or notebook per student.
  • Every answer should include code, output, and a short explanation.
  • At Levels 4 and 5, students must write what pattern they observed.
  • Bonus marks for elegant vectorized solutions.

Level 1 — Ground zero

  • Create a NumPy array containing numbers from 1 to 20.
  • Print only even numbers.
  • Print the first 5 elements, last 5 elements, and every second element.
  • Write one sentence: how is this different from using a Python list?

Level 2 — Basic control of arrays

  • Create a 3 × 4 matrix with values from 1 to 12.
  • Print the shape, size, and number of dimensions.
  • Print the second row and third column.
  • Reshape the same values into a 2 × 6 matrix.
  • Explain why reshape works here.

Level 3 — Vectorization and aggregation

  • Create an array of marks for 10 students.
  • Add 5 grace marks to all students using vectorized code.
  • Find total, average, highest, lowest, median, and standard deviation.
  • Count how many students scored above the class average.
  • Sort the marks in ascending order.

Level 4 — Broadcasting and axis work

  • Create a 4 × 3 matrix representing 4 students and 3 subjects.
  • Find row sums and column sums.
  • Add subject-wise bonus marks using broadcasting, for example [2, 5, 3].
  • Find each student’s maximum score.
  • Find the average score in each subject.
  • Explain what axis=0 and axis=1 mean in your own words.

Level 5 — Expert point 100

  • Create a random dataset of 50 student marks using np.random.randint(30, 100, 50).
  • Reshape it into a 10 × 5 matrix.
  • Interpret rows as students and columns as tests.
  • Find each student’s total and average.
  • Find the topper based on total marks.
  • Find how many test scores are below 40.
  • Create a 10 × 10 random integer matrix and normalize it by dividing by 100.
  • Write 5–8 lines explaining how these kinds of operations prepare data for AI-ML.

Challenge tasks

  • Compare list addition and array addition with your own example.
  • Create a timing comparison using Python time or timeit for loop style versus vectorized style.
  • Transpose a matrix and describe what changed.
  • Show one reshape attempt that fails and explain why it fails.

Mini project

Build a Student Dataset Analyzer. The program should create or accept marks, summarize them, reshape them into a readable matrix, identify topper and below-average students, and print a final interpretation paragraph. 

import numpy as np

marks = np.random.randint(35, 100, 20)
print("Marks:", marks)
print("Average:", np.mean(marks))
print("Top score:", np.max(marks))
print("Below average:", np.sum(marks < np.mean(marks)))

Evaluation rubric

Criterion Marks focus What strong work looks like
Correctness 25 Code runs and gives correct outputs
Concept understanding 25 Student can explain shape, slicing, vectorization, and broadcasting clearly
Clean coding 15 Readable code, sensible variable names, useful comments
Interpretation 20 Student explains what the results mean, not only what they are
Depth / challenge attempt 15 Student goes beyond minimum and attempts higher-level tasks