np.count_nonzero() for counting in NumPy - Time & Space Complexity

Time Complexity: np.count_nonzero() for counting

O(n)

Understanding Time Complexity

We want to understand how the time to count non-zero elements grows as the array gets bigger.

How does the work change when the input size increases?

Scenario Under Consideration

Analyze the time complexity of the following code snippet.

import numpy as np

arr = np.random.randint(0, 5, size=1000)
count = np.count_nonzero(arr)

This code creates an array of 1000 numbers and counts how many are not zero.

Identify Repeating Operations

• Primary operation: Checking each element to see if it is non-zero.
• How many times: Once for every element in the array.

How Execution Grows With Input

As the array size grows, the number of checks grows at the same rate.

Input Size (n)	Approx. Operations
10	10 checks
100	100 checks
1000	1000 checks

Pattern observation: The work grows directly with the number of elements.

Final Time Complexity

Time Complexity: O(n)

This means the time to count non-zero elements grows in a straight line with the array size.

Common Mistake

[X] Wrong: "Counting non-zero elements is instant no matter the array size."

[OK] Correct: The function must look at each element once, so bigger arrays take more time.

Interview Connect

Knowing how counting operations scale helps you explain efficiency clearly and shows you understand how data size affects performance.

Self-Check

"What if we count non-zero elements only in a slice of the array? How would the time complexity change?"

Practice

1.

What does the np.count_nonzero() function do in NumPy?

easy

A. Calculates the sum of all elements in an array

B. Returns the shape of the array

C. Finds the maximum value in an array

D. Counts how many elements in an array are not zero

Solution

1. Step 1: Understand the function purpose

   np.count_nonzero() counts elements that are not zero in the array.

2. Step 2: Compare with other options

   Other options describe different functions like sum, max, or shape, which are not what np.count_nonzero() does.

3. Final Answer:

   Counts how many elements in an array are not zero -> Option D

4. Quick Check:

   Counting non-zero elements = Counts how many elements are not zero [OK]

Hint: Remember: count_nonzero counts non-zero values only [OK]

Common Mistakes:

• Confusing count_nonzero with sum or max functions
• Thinking it returns the array shape
• Assuming it counts zero elements

2.

Which of the following is the correct syntax to count non-zero elements in a NumPy array arr?

arr = np.array([1, 0, 3, 0, 5])

easy

A. np.count_nonzero = arr

B. np.count_nonzero(arr)

C. arr.count_nonzero()

D. np.count(arr != 0)

Solution

1. Step 1: Identify correct function usage

   The function np.count_nonzero() is called with the array as argument: np.count_nonzero(arr).

2. Step 2: Check other options for errors

   np.count_nonzero = arr tries to assign instead of call; arr.count_nonzero() uses method not available on array; np.count(arr != 0) uses a non-existent function np.count().

3. Final Answer:

   np.count_nonzero(arr) -> Option B

4. Quick Check:

   Correct syntax is np.count_nonzero(array) [OK]

Hint: Use np.count_nonzero(array) to count non-zero values [OK]

Common Mistakes:

• Using assignment instead of function call
• Calling count_nonzero as a method on array
• Using non-existent np.count function

3.

What is the output of the following code?

import numpy as np
arr = np.array([[0, 1, 2], [3, 0, 0], [4, 5, 6]])
count = np.count_nonzero(arr, axis=0)
print(count)

medium

A. [3 2 3]

B. [3 3 3]

C. [2 3 2]

D. [2 2 2]

Solution

1. Step 1: Understand axis=0 counting

   Counting non-zero elements along columns (axis=0) means counting down each column.

2. Step 2: Count non-zero per column

   Column 1: values [0,3,4] -> non-zero count = 2 (3 and 4)
   Column 2: values [1,0,5] -> non-zero count = 2 (1 and 5)
   Column 3: values [2,0,6] -> non-zero count = 2 (2 and 6)

3. Final Answer:

   [3 2 3] -> Option A

4. Quick Check:

   Count non-zero per column = [3 2 3] [OK]

Hint: axis=0 counts down columns, axis=1 counts across rows [OK]

Common Mistakes:

• Counting zeros instead of non-zero
• Confusing axis=0 with axis=1
• Miscounting elements per column

4.

Find the error in this code snippet and choose the correct fix:

import numpy as np
arr = np.array([1, 0, 2, 0, 3])
count = np.count_nonzero(arr, axis=1)
print(count)

medium

A. Remove axis=1 because arr is 1D, use np.count_nonzero(arr) instead

B. Change axis=1 to axis=0 to fix the error

C. Use arr.count_nonzero() method instead

D. No error, code runs fine

Solution

1. Step 1: Identify array dimension

   Array arr is 1D, so axis=1 is invalid (no second axis).

2. Step 2: Correct function call

   Remove axis argument to count all non-zero elements: np.count_nonzero(arr).

3. Final Answer:

   Remove axis=1 because arr is 1D, use np.count_nonzero(arr) instead -> Option A

4. Quick Check:

   1D arrays have no axis=1, so omit axis [OK]

Hint: Check array shape before using axis in count_nonzero [OK]

Common Mistakes:

• Using axis=1 on 1D arrays causes errors
• Trying to call count_nonzero as array method
• Assuming axis=0 fixes all axis errors

5.

You have a 2D NumPy array representing attendance (1 for present, 0 for absent) of 4 students over 5 days:

attendance = np.array([
  [1, 0, 1, 1, 0],
  [0, 0, 1, 0, 0],
  [1, 1, 1, 1, 1],
  [0, 0, 0, 0, 0]
])

Which code correctly counts how many days each student was present?

hard

A. np.sum(attendance, axis=0)

B. np.count_nonzero(attendance, axis=0)

C. np.count_nonzero(attendance, axis=1)

D. np.count_nonzero(attendance)

Solution

1. Step 1: Understand data layout

   Rows represent students, columns represent days. Counting days present per student means counting non-zero per row (axis=1).

2. Step 2: Choose correct axis

   Use np.count_nonzero(attendance, axis=1) to count non-zero values per student (row).

3. Final Answer:

   np.count_nonzero(attendance, axis=1) -> Option C

4. Quick Check:

   Count per row (student) = axis=1 [OK]

Hint: Count per student = count_nonzero with axis=1 [OK]

Common Mistakes:

• Using axis=0 counts per day, not per student
• Using np.sum instead of count_nonzero (works but different function)
• Counting total non-zero without axis