Record arrays in NumPy - Time & Space Complexity

Time Complexity: Record arrays

O(n)

Understanding Time Complexity

We want to understand how the time to access and manipulate data in numpy record arrays changes as the data size grows.

Specifically, how does the cost grow when working with structured data stored in record arrays?

Scenario Under Consideration

Analyze the time complexity of the following code snippet.

import numpy as np

# Define a record array with 3 fields
data = np.recarray(1000, dtype=[('name', 'U10'), ('age', 'i4'), ('score', 'f4')])

# Access the 'age' field for all records
ages = data.age

# Compute the average age
average_age = np.mean(ages)

This code creates a record array with 1000 entries, accesses one field for all records, and calculates the average of that field.

Identify Repeating Operations

Identify the loops, recursion, array traversals that repeat.

• Primary operation: Accessing the 'age' field for all 1000 records and computing the mean.
• How many times: The operation touches each record once, so 1000 times.

How Execution Grows With Input

As the number of records grows, the time to access and process each record grows proportionally.

Input Size (n)	Approx. Operations
10	About 10 operations to access and process
100	About 100 operations
1000	About 1000 operations

Pattern observation: The operations grow linearly with the number of records.

Final Time Complexity

Time Complexity: O(n)

This means the time to access and compute over the record array grows directly in proportion to the number of records.

Common Mistake

[X] Wrong: "Accessing a field in a record array is instant and does not depend on the number of records."

[OK] Correct: Even though the field access looks simple, numpy must read each record's field, so the time grows with the number of records.

Interview Connect

Understanding how structured data access scales helps you reason about performance in real data tasks and shows you can analyze array operations clearly.

Self-Check

"What if we accessed multiple fields at once instead of just one? How would the time complexity change?"

Practice

1. What is the main advantage of using a record array in numpy?

easy

A. It speeds up numerical calculations on large arrays.

B. It automatically sorts data based on values.

C. It allows storing different data types in one array with named fields.

D. It compresses data to save memory.

Solution

1. Step 1: Understand record arrays

   Record arrays let you store mixed data types in one numpy array by using named fields.

2. Step 2: Compare options

   Only It allows storing different data types in one array with named fields. correctly describes this feature. Others describe unrelated features.

3. Final Answer:

   It allows storing different data types in one array with named fields. -> Option C

4. Quick Check:

   Record arrays = mixed types + named fields [OK]

Hint: Remember: record arrays hold mixed types with names [OK]

Common Mistakes:

• Confusing record arrays with regular numeric arrays
• Thinking record arrays sort data automatically
• Assuming record arrays compress data

2. Which of the following is the correct way to create a numpy record array with fields 'name' (string) and 'age' (integer)?

easy

A. np.rec.array([("Alice", 25), ("Bob", 30)], dtype=[('name', 'U10'), ('age', 'i4')])

B. np.array([("Alice", 25), ("Bob", 30)], dtype=[('name', 'i4'), ('age', 'U10')])

C. np.rec.array(["Alice", 25, "Bob", 30], dtype=[('name', 'U10'), ('age', 'i4')])

D. np.rec.array([(25, "Alice"), (30, "Bob")], dtype=[('name', 'U10'), ('age', 'i4')])

Solution

1. Step 1: Check data and dtype matching

   np.rec.array([("Alice", 25), ("Bob", 30)], dtype=[('name', 'U10'), ('age', 'i4')]) matches tuples of (string, int) with dtype [('name', 'U10'), ('age', 'i4')].

2. Step 2: Validate other options

   np.array([("Alice", 25), ("Bob", 30)], dtype=[('name', 'i4'), ('age', 'U10')]) swaps types incorrectly; C has wrong input format; A swaps field order.

3. Final Answer:

   np.rec.array([("Alice", 25), ("Bob", 30)], dtype=[('name', 'U10'), ('age', 'i4')]) -> Option A

4. Quick Check:

   Data matches dtype order and types [OK]

Hint: Match tuple order with dtype fields exactly [OK]

Common Mistakes:

• Swapping field order between data and dtype
• Using wrong data types in dtype
• Passing flat list instead of list of tuples

3. What will be the output of the following code?

import numpy as np
rec = np.rec.array([(1, 2.5), (3, 4.5)], dtype=[('x', 'i4'), ('y', 'f4')])
print(rec.x + rec.y)

medium

A. TypeError

B. [3 7]

C. [1 3]

D. [3.5 7.5]

Solution

1. Step 1: Understand data and fields

   rec.x is integer array [1, 3], rec.y is float array [2.5, 4.5].

2. Step 2: Add integer and float arrays element-wise

   Adding [1, 3] + [2.5, 4.5] results in [3.5, 7.5] as floats.

3. Final Answer:

   [3.5 7.5] -> Option D

4. Quick Check:

   1+2.5=3.5 and 3+4.5=7.5 [OK]

Hint: Adding int and float fields results in float array [OK]

Common Mistakes:

• Expecting integer output instead of float
• Confusing field names or types
• Thinking addition causes error

4. Identify the error in this code snippet:

import numpy as np
rec = np.rec.array([(1, 'a'), (2, 'b')], dtype=[('num', 'i4'), ('char', 'U1')])
print(rec.num + rec.char)

medium

A. You cannot add integer and string fields directly.

B. The dtype specification is incorrect.

C. The data tuples have wrong length.

D. The record array must be created with np.array, not np.rec.array.

Solution

1. Step 1: Analyze the operation

   rec.num is integer array, rec.char is string array.

2. Step 2: Check addition of int and string

   Adding int + string causes a TypeError in numpy.

3. Final Answer:

   You cannot add integer and string fields directly. -> Option A

4. Quick Check:

   int + string = TypeError [OK]

Hint: Cannot add numbers and strings directly in numpy [OK]

Common Mistakes:

• Assuming dtype is wrong instead of operation
• Thinking np.rec.array is incorrect here
• Ignoring type mismatch in addition

5. You have a numpy record array rec with fields 'id' (int), 'score' (float), and 'passed' (bool). How do you create a new record array containing only records where passed is True and score is above 80?

hard

A. rec[rec.passed or rec.score > 80]

B. rec[(rec.passed) & (rec.score > 80)]

C. rec[rec.passed and rec.score > 80]

D. rec[(rec.passed) | (rec.score > 80)]

Solution

1. Step 1: Understand filtering syntax

   Use boolean indexing with & for element-wise AND, parentheses needed.

2. Step 2: Evaluate options

   rec[(rec.passed) & (rec.score > 80)] correctly uses (rec.passed) & (rec.score > 80). Options B and C use Python 'or'/'and' which don't work element-wise. rec[(rec.passed) | (rec.score > 80)] uses | (OR) instead of AND.

3. Final Answer:

   rec[(rec.passed) & (rec.score > 80)] -> Option B

4. Quick Check:

   Use & with parentheses for element-wise AND [OK]

Hint: Use & with parentheses for element-wise conditions [OK]

Common Mistakes:

• Using 'and' or 'or' instead of '&' or '|' for arrays
• Forgetting parentheses around conditions
• Using | instead of & for AND condition