Record arrays in NumPy - Step-by-Step Execution

Concept Flow - Record arrays

Define dtype with field names and types

↓

Create record array with structured data

↓

Access fields by name

↓

Perform operations on fields

↓

Use record array for analysis or display

Record arrays store data with named fields, like a table with columns. You define the structure, create the array, then access fields by name for easy analysis.

Execution Sample

NumPy

import numpy as np

dtype = [('name', 'U10'), ('age', 'i4'), ('score', 'f4')]
data = [('Alice', 25, 88.5), ('Bob', 30, 92.0), ('Cathy', 22, 79.5)]
rec_arr = np.array(data, dtype=dtype)
print(rec_arr['age'])

Creates a record array with fields name, age, score and prints the age column.

Execution Table

Step	Action	Data State	Field Access	Output
1	Define dtype with fields	[('name', 'U10'), ('age', 'i4'), ('score', 'f4')]	N/A	N/A
2	Create record array with data	[('Alice', 25, 88.5), ('Bob', 30, 92.0), ('Cathy', 22, 79.5)]	N/A	N/A
3	Access 'age' field	Record array unchanged	rec_arr['age']	[25 30 22]
4	Access 'name' field	Record array unchanged	rec_arr['name']	['Alice' 'Bob' 'Cathy']
5	Access 'score' field	Record array unchanged	rec_arr['score']	[88.5 92. 79.5]
6	Filter ages > 23	Record array unchanged	rec_arr['age'] > 23	[ True True False]
7	Select records with age > 23	Record array unchanged	rec_arr[rec_arr['age'] > 23]	[('Alice', 25, 88.5) ('Bob', 30, 92.0)]
8	Exit	End of operations	N/A	N/A

💡 All steps executed; final filtered record array shown.

Variable Tracker

Variable	Start	After Step 2	After Step 7	Final
dtype	undefined	[('name', 'U10'), ('age', 'i4'), ('score', 'f4')]	[('name', 'U10'), ('age', 'i4'), ('score', 'f4')]	[('name', 'U10'), ('age', 'i4'), ('score', 'f4')]
data	undefined	[('Alice', 25, 88.5), ('Bob', 30, 92.0), ('Cathy', 22, 79.5)]	[('Alice', 25, 88.5), ('Bob', 30, 92.0), ('Cathy', 22, 79.5)]	[('Alice', 25, 88.5), ('Bob', 30, 92.0), ('Cathy', 22, 79.5)]
rec_arr	undefined	[('Alice', 25, 88.5), ('Bob', 30, 92.0), ('Cathy', 22, 79.5)]	[('Alice', 25, 88.5), ('Bob', 30, 92.0), ('Cathy', 22, 79.5)]	[('Alice', 25, 88.5), ('Bob', 30, 92.0), ('Cathy', 22, 79.5)]

Key Moments - 3 Insights

Why can we access fields by name like rec_arr['age']?

Because the record array is created with a dtype that defines named fields, numpy allows direct access to each field as if it were a column. See execution_table step 3.

What happens when we filter with rec_arr['age'] > 23?

This creates a boolean array showing which records meet the condition. We then use it to select only those records. See execution_table steps 6 and 7.

Is the original record array changed after filtering?

No, filtering returns a new array with selected records. The original record array stays the same. See execution_table step 7.

Visual Quiz - 3 Questions

Look at the execution table, what is the output of rec_arr['score'] at step 5?

A [88.5 92. 79.5]

B [25 30 22]

C ['Alice' 'Bob' 'Cathy']

D [True True False]

💡 Hint

Check the 'Output' column in execution_table row 5.

At which step does the condition rec_arr['age'] > 23 become False for some records?

A Step 3

B Step 7

C Step 6

D Step 4

💡 Hint

Look at the 'Field Access' and 'Output' columns in execution_table row 6.

If we change the dtype to remove the 'score' field, what happens when accessing rec_arr['score']?

A It returns an empty array.

B It raises an IndexError.

C It returns all zeros.

D It returns the 'age' field instead.

💡 Hint

Accessing a non-existent field in a record array causes an error.

Concept Snapshot

Record arrays store data with named fields.
Define dtype with field names and types.
Create array with structured data.
Access fields by name like columns.
Filter and select records easily.
Useful for table-like data in numpy.

Full Transcript

Record arrays in numpy let you store data with named fields, like columns in a table. First, you define a dtype that lists each field's name and data type. Then you create a numpy array with this dtype and your data. You can access each field by its name, for example rec_arr['age'] gives all ages. You can also filter records by conditions on fields, like selecting only those with age greater than 23. The original array stays unchanged when filtering; a new array is returned. This makes record arrays very handy for structured data analysis.