Combining fancy and slice indexing in NumPy - Time & Space Complexity

Time Complexity: Combining fancy and slice indexing

O(k)

Understanding Time Complexity

We want to understand how the time needed changes when we combine fancy and slice indexing in numpy arrays.

How does the work grow when we pick many elements using both methods together?

Scenario Under Consideration

Analyze the time complexity of the following code snippet.

import numpy as np

arr = np.arange(1000)
indices = [2, 5, 7, 10]
result = arr[indices[1:3]]

This code picks elements from a numpy array using a slice of a list of indices (fancy indexing combined with slicing).

Identify Repeating Operations

Identify the loops, recursion, array traversals that repeat.

• Primary operation: Accessing elements at selected indices from the array.
• How many times: Once for each index in the slice (here 2 times for indices 5 and 7).

How Execution Grows With Input

As the number of indices selected grows, the work grows proportionally.

Input Size (n)	Approx. Operations
10	10 element accesses
100	100 element accesses
1000	1000 element accesses

Pattern observation: The time grows linearly with the number of indices selected.

Final Time Complexity

Time Complexity: O(k)

This means the time grows directly with how many elements you pick using the combined indexing.

Common Mistake

[X] Wrong: "Combining fancy and slice indexing makes the operation take time proportional to the whole array size."

[OK] Correct: Actually, numpy only accesses the elements you ask for, so time depends on how many indices you select, not the full array size.

Interview Connect

Understanding how indexing time grows helps you write efficient data selection code and explain your choices clearly in real projects or interviews.

Self-Check

"What if we replaced the slice with a full list of indices? How would the time complexity change?"

Practice

1. What does combining fancy indexing and slice indexing in NumPy allow you to do?
arr[[0, 2], 1:4] selects:

easy

A. Rows 0 to 2, columns 1 and 4

B. Rows 1 to 4, columns 0 and 2

C. Rows 0 and 2, columns 1 to 3

D. Rows 1 and 4, columns 0 to 2

Solution

1. Step 1: Understand fancy indexing for rows

   The fancy index [0, 2] selects rows 0 and 2 specifically.

2. Step 2: Understand slice indexing for columns

   The slice 1:4 selects columns starting at 1 up to but not including 4, so columns 1, 2, and 3.

3. Final Answer:

   Rows 0 and 2, columns 1 to 3 -> Option C

4. Quick Check:

   Fancy rows + slice columns = selected subarray [OK]

Hint: Fancy picks rows, slice picks continuous columns [OK]

Common Mistakes:

• Confusing rows and columns in indexing
• Forgetting slice end is exclusive
• Mixing up fancy and slice roles

2. Which of the following is the correct syntax to select rows 1 and 3 and columns 2 to 4 from a NumPy array arr?

easy

A. arr[[1, 3], 2:5]

B. arr[1:3, 2:5]

C. arr[1, 3, 2:5]

D. arr[[1, 3], [2:5]]

Solution

1. Step 1: Identify fancy indexing for rows

   To select rows 1 and 3, use fancy indexing with a list: [1, 3].

2. Step 2: Use slice for columns

   To select columns 2 to 4, use slice 2:5 (end exclusive).

3. Final Answer:

   arr[[1, 3], 2:5] -> Option A

4. Quick Check:

   Fancy rows + slice columns syntax correct [OK]

Hint: Fancy indexing uses lists, slice uses colon syntax [OK]

Common Mistakes:

• Using commas inside fancy index list
• Using slice inside list brackets
• Incorrect slice end index

3. Given the array:

import numpy as np
arr = np.array([[10, 20, 30, 40],
                [50, 60, 70, 80],
                [90, 100, 110, 120],
                [130, 140, 150, 160]])
print(arr[[1, 3], 1:3])

What is the output?

medium

A. [[ 70 80]
[150 160]]

B. [[ 60 70]
[140 150]]

C. [[ 50 60]
[130 140]]

D. [[ 60 70 80]
[140 150 160]]

Solution

1. Step 1: Select rows using fancy indexing

   Rows 1 and 3 are:
   [50, 60, 70, 80]
   [130, 140, 150, 160]

2. Step 2: Select columns 1 to 2 (slice 1:3)

   Columns 1 and 2 for these rows are:
   Row 1: 60, 70
   Row 3: 140, 150

3. Final Answer:

   [[ 60 70] [140 150]] -> Option B

4. Quick Check:

   Fancy rows + slice columns output matches [OK]

Hint: Pick rows first, then slice columns [OK]

Common Mistakes:

• Selecting wrong rows or columns
• Including column 3 by mistake
• Confusing row and column indices

4. What is wrong with this code snippet?

import numpy as np
arr = np.arange(16).reshape(4,4)
result = arr[1:3, [0, 2]]
print(result)

medium

A. It prints rows 1 and 2 with columns 0 and 2 correctly

B. It raises an error because slice and fancy indexing cannot be combined

C. It prints rows 0 and 2 with columns 1 and 3 incorrectly

D. It raises an error because the fancy index list is invalid

Solution

1. Step 1: Understand slice and fancy indexing combination

   Slice 1:3 selects rows 1 and 2. Fancy index [0, 2] selects columns 0 and 2.

2. Step 2: Check if code runs and output

   This combination is valid in NumPy and prints the selected subarray without error.

3. Final Answer:

   It prints rows 1 and 2 with columns 0 and 2 correctly -> Option A

4. Quick Check:

   Slice rows + fancy columns works fine [OK]

Hint: Slice and fancy indexing can combine on different axes [OK]

Common Mistakes:

• Thinking slice and fancy indexing cannot mix
• Confusing which axis uses slice or fancy
• Expecting an error when code is valid

5. You have a 5x5 NumPy array arr. You want to select rows 0, 2, and 3, but only columns 1 to 3 inclusive. Which code correctly does this?

hard

A. arr[1:4, [0, 2, 4]]

B. arr[0:5:2, 1:4]

C. arr[[0, 2, 4], [1, 2, 3]]

D. arr[[0, 2, 3], 1:4]

Solution

1. Step 1: Identify fancy indexing for rows

   Rows 0, 2, and 3 are selected using fancy indexing: [0, 2, 3].

2. Step 2: Use slice for columns 1 to 3 inclusive

   Slice 1:4 selects columns 1, 2, and 3 (end index exclusive).

3. Final Answer:

   arr[[0, 2, 3], 1:4] -> Option D

4. Quick Check:

   Fancy rows + slice columns selects correct subarray [OK]

Hint: Fancy list for rows, slice for continuous columns [OK]

Common Mistakes:

• Using slice for rows instead of fancy list
• Using fancy list for columns incorrectly
• Off-by-one error in slice end index