np.split() for dividing arrays in NumPy - Time & Space Complexity

Time Complexity: np.split() for dividing arrays

O(n)

Understanding Time Complexity

We want to understand how the time needed to split an array changes as the array gets bigger.

Specifically, how does dividing an array into parts affect the work done?

Scenario Under Consideration

Analyze the time complexity of the following code snippet.

import numpy as np

arr = np.arange(1000)
splits = np.split(arr, 10)

This code creates an array of 1000 numbers and splits it into 10 equal parts.

Identify Repeating Operations

Identify the loops, recursion, array traversals that repeat.

• Primary operation: Copying parts of the original array into new sub-arrays.
• How many times: The array is split into 10 parts, so this copying happens 10 times.

How Execution Grows With Input

As the array size grows, the total amount of data copied grows proportionally.

Input Size (n)	Approx. Operations
10	About 10 copies of 1 element each = 10 operations
100	About 10 copies of 10 elements each = 100 operations
1000	About 10 copies of 100 elements each = 1000 operations

Pattern observation: The total work grows roughly in direct proportion to the size of the input array.

Final Time Complexity

Time Complexity: O(n)

This means the time to split the array grows linearly with the size of the array.

Common Mistake

[X] Wrong: "Splitting an array into parts is very fast and does not depend on the array size."

[OK] Correct: Actually, splitting involves copying data into new arrays, so the bigger the array, the more copying is needed, which takes more time.

Interview Connect

Understanding how array operations scale helps you explain your code choices clearly and shows you know how to handle data efficiently.

Self-Check

"What if we used views instead of copies when splitting? How would the time complexity change?"

Practice

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

easy

A. It merges multiple arrays into one large array.

B. It divides an array into multiple smaller arrays at specified indices or into equal parts.

C. It sorts the elements of an array in ascending order.

D. It reshapes an array into a different shape without changing data.

Solution

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

   The function is designed to break a large array into smaller arrays either by specifying split points or by dividing into equal parts.

2. Step 2: Compare with other array functions

   Unlike merging, sorting, or reshaping, np.split specifically divides arrays.

3. Final Answer:

   It divides an array into multiple smaller arrays at specified indices or into equal parts. -> Option B

4. Quick Check:

   np.split() = divides array [OK]

Hint: Remember: split means break into parts [OK]

Common Mistakes:

• Confusing split with merge
• Thinking it sorts arrays
• Mixing with reshape

2. Which of the following is the correct syntax to split a 1D NumPy array arr into 3 equal parts using np.split()?

easy

A. np.split(arr, 3)

B. np.split(arr, [3])

C. np.split(arr, indices=3)

D. np.split(arr, sections=3)

Solution

1. Step 1: Recall np.split() syntax

   The function signature is np.split(array, indices_or_sections), where indices_or_sections can be an integer for equal parts.

2. Step 2: Identify correct usage

   Passing 3 as the second argument splits the array into 3 equal parts. Keyword arguments like 'indices_or_sections' are not used explicitly in common syntax.

3. Final Answer:

   np.split(arr, 3) -> Option A

4. Quick Check:

   np.split(arr, 3) splits into 3 parts [OK]

Hint: Use integer directly for equal parts [OK]

Common Mistakes:

• Using list instead of integer for equal parts
• Wrong keyword arguments
• Confusing indices with number of parts

3. What is the output of the following code?

import numpy as np
arr = np.array([10, 20, 30, 40, 50, 60])
splits = np.split(arr, [2, 4])
print(splits)

medium

A. [array([10]), array([20, 30, 40]), array([50, 60])]

B. [array([10, 20, 30]), array([40, 50, 60])]

C. [array([10, 20]), array([30, 40]), array([50, 60])]

D. [array([10, 20]), array([30]), array([40, 50, 60])]

Solution

1. Step 1: Understand split indices

   The list [2, 4] means split at index 2 and then at index 4.

2. Step 2: Divide the array accordingly

   arr[:2] = [10, 20], arr[2:4] = [30, 40], arr[4:] = [50, 60]

3. Final Answer:

   [array([10, 20]), array([30, 40]), array([50, 60])] -> Option C

4. Quick Check:

   Split at 2 and 4 gives 3 parts [OK]

Hint: Split indices mark where to cut array [OK]

Common Mistakes:

• Misunderstanding split points
• Off-by-one errors in slicing
• Confusing output format

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

import numpy as np
arr = np.array([1, 2, 3, 4, 5])
splits = np.split(arr, 3)
print(splits)

medium

A. The array length is not divisible by 3, so equal splits are impossible.

B. np.split() requires a list of indices, not an integer.

C. The array must be 2D to use np.split().

D. The print statement syntax is incorrect.

Solution

1. Step 1: Check array length and split parts

   The array length is 5, but we want to split into 3 equal parts.

2. Step 2: Understand np.split() behavior

   np.split() requires the array length to be divisible by the number of parts when splitting equally, otherwise it raises an error.

3. Final Answer:

   The array length is not divisible by 3, so equal splits are impossible. -> Option A

4. Quick Check:

   Length 5 not divisible by 3 causes error [OK]

Hint: Check divisibility of array length by parts [OK]

Common Mistakes:

• Using integer splits on incompatible length
• Confusing np.split with np.array_split
• Assuming print error

5. You have a 2D NumPy array data with shape (6, 4). You want to split it into 3 equal parts along the first axis (rows) for separate analysis. Which code correctly does this using np.split()?

hard

A. np.split(data, [1, 3, 5], axis=1)

B. np.split(data, 3, axis=1)

C. np.split(data, [2, 4], axis=1)

D. np.split(data, 3, axis=0)

Solution

1. Step 1: Understand array shape and axis

   data has 6 rows and 4 columns. Splitting into 3 equal parts along rows means each part has 2 rows.

2. Step 2: Use np.split() with correct axis and parts

   np.split(data, 3, axis=0) splits rows into 3 equal parts. axis=1 would split columns, which is not desired.

3. Final Answer:

   np.split(data, 3, axis=0) -> Option D

4. Quick Check:

   Split rows into 3 parts with axis=0 [OK]

Hint: Use axis=0 to split rows, axis=1 for columns [OK]

Common Mistakes:

• Splitting along wrong axis
• Using indices instead of number for equal parts
• Ignoring array shape