np.unique() for unique values in NumPy - Time & Space Complexity

Time Complexity: np.unique() for unique values

O(n log n)

Understanding Time Complexity

We want to understand how the time needed to find unique values in an array changes as the array gets bigger.

How does the work grow when we ask numpy to find unique items?

Scenario Under Consideration

Analyze the time complexity of the following code snippet.

import numpy as np

arr = np.array([3, 1, 2, 3, 4, 1, 5])
unique_vals = np.unique(arr)
print(unique_vals)

This code finds all unique values in the array arr and returns them sorted.

Identify Repeating Operations

Identify the loops, recursion, array traversals that repeat.

• Primary operation: Sorting the array elements to group duplicates.
• How many times: The sorting process compares elements multiple times, roughly proportional to the number of elements times the logarithm of that number.

How Execution Grows With Input

As the array size grows, the time to find unique values grows a bit faster than the size itself but not as fast as the square of the size.

Input Size (n)	Approx. Operations
10	About 30 to 40 operations
100	About 600 to 700 operations
1000	About 10,000 to 12,000 operations

Pattern observation: The operations grow faster than the input size but slower than its square, roughly like size times log of size.

Final Time Complexity

Time Complexity: O(n log n)

This means the time to find unique values grows a bit faster than the number of items but not as fast as checking every pair.

Common Mistake

[X] Wrong: "Finding unique values takes the same time no matter how many items there are."

[OK] Correct: The process needs to compare and sort items, so more items mean more work, not a fixed time.

Interview Connect

Understanding how numpy finds unique values helps you explain how data processing scales, a useful skill when working with real datasets.

Self-Check

"What if the input array was already sorted? How would the time complexity change?"

Practice

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

easy

A. Calculates the sum of all elements

B. Sorts the array in descending order

C. Reverses the order of elements

D. Finds all unique values in an array

Solution

1. Step 1: Understand the purpose of np.unique()

   This function is designed to find all different (unique) values in a NumPy array or list.

2. Step 2: Compare with other options

   Sorting, summing, or reversing are different operations and not what np.unique() does.

3. Final Answer:

   Finds all unique values in an array -> Option D

4. Quick Check:

   np.unique() = unique values [OK]

Hint: Remember: unique means different values only [OK]

Common Mistakes:

• Confusing unique with sorting
• Thinking it sums values
• Assuming it reverses array

2. Which of the following is the correct syntax to get unique values from a NumPy array arr?

easy

A. np.arr.unique()

B. np.unique(arr)

C. unique(arr)

D. arr.unique()

Solution

1. Step 1: Recall the correct function call

   The function unique is part of the NumPy module and is called as np.unique().

2. Step 2: Check other options for errors

   arr.unique() is not a NumPy array method, unique(arr) misses the module prefix, and np.arr.unique() is invalid syntax.

3. Final Answer:

   np.unique(arr) -> Option B

4. Quick Check:

   Correct syntax = np.unique(arr) [OK]

Hint: Use np.unique(array) always [OK]

Common Mistakes:

• Calling unique as method on array
• Missing np prefix
• Wrong module usage

3. What is the output of this code?

import numpy as np
arr = np.array([3, 1, 2, 3, 2, 1, 4])
print(np.unique(arr))

medium

A. [1 2 3 4]

B. [3 1 2 4]

C. [4 3 2 1]

D. [1 1 2 2 3 3 4]

Solution

1. Step 1: Identify unique values in the array

   The array contains values 3, 1, 2, 3, 2, 1, 4. Unique values are 1, 2, 3, and 4.

2. Step 2: Understand np.unique() output order

   np.unique() returns sorted unique values, so output is [1 2 3 4].

3. Final Answer:

   [1 2 3 4] -> Option A

4. Quick Check:

   Unique sorted values = [1 2 3 4] [OK]

Hint: Unique values are sorted by default [OK]

Common Mistakes:

• Ignoring sorting order
• Listing duplicates
• Wrong output format

4. The following code throws an error. What is the problem?

import numpy as np
arr = [1, 2, 2, 3]
print(np.unique(arr, return_counts=True))

medium

A. Incorrect argument name, should be count_return=True

B. List input causes error, must convert to np.array first

C. No error, code runs fine

D. np.unique() does not support return_counts argument

Solution

1. Step 1: Check input type compatibility

   np.unique() accepts lists or arrays as input without error.

2. Step 2: Verify return_counts argument

   The argument return_counts=True is valid and returns counts of unique values.

3. Final Answer:

   No error, code runs fine -> Option C

4. Quick Check:

   List input + return_counts works [OK]

Hint: np.unique accepts lists and arrays [OK]

Common Mistakes:

• Thinking list input causes error
• Wrong argument name
• Assuming return_counts unsupported

5. You have a 2D NumPy array:

arr = np.array([[1, 2, 2], [3, 1, 4]])

How do you get all unique values from this 2D array as a sorted 1D array?

hard

A. np.unique(arr)

B. np.unique(arr, axis=1)

C. np.unique(arr, axis=0)

D. arr.unique()

Solution

1. Step 1: Understand np.unique() on 2D arrays

   Calling np.unique() without axis flattens the array and returns unique sorted values.

2. Step 2: Check axis arguments

   Using axis=0 or axis=1 returns unique rows or columns, not unique elements overall.

3. Final Answer:

   np.unique(arr) -> Option A

4. Quick Check:

   Flatten and unique = np.unique(arr) [OK]

Hint: No axis means flatten and find unique [OK]

Common Mistakes:

• Using axis to get unique elements
• Calling unique as method
• Expecting 2D unique output