NumPy Field Access – Structured Arrays and Column Access Explained in Python

🐍 NumPy – Field Access

In NumPy, field access is used when working with structured arrays (also called record arrays).

Unlike normal NumPy arrays, structured arrays allow you to assign names to columns, similar to a database table or pandas DataFrame.

Field access allows you to:

• Access data using column names
• Work with structured datasets
• Store mixed data types
• Handle tabular data efficiently

---

What is Field Access in NumPy?

Field access means retrieving values from a structured array using named fields (column names) instead of numeric indices.

---

Example Concept

Name     Age     Score
------------------------
Aman     20      85
John     22      90
Sara     21      88

You can access:

• Name column
• Age column
• Score column

---

🟢 Creating Structured Arrays

To use field access, we must define a data type with named fields.

Example

import numpy as np

data = np.array([
    ("Aman", 20, 85),
    ("John", 22, 90),
    ("Sara", 21, 88)
], dtype=[("name", "U10"), ("age", "i4"), ("score", "i4")])

print(data)

---

Output:

[('Aman', 20, 85)
 ('John', 22, 90)
 ('Sara', 21, 88)]

---

🔵 Accessing Fields (Column Access)

You can access data using field names.

---

Access Names Column

print(data["name"])

Output:

['Aman' 'John' 'Sara']

---

Access Age Column

print(data["age"])

Output:

[20 22 21]

---

Access Score Column

print(data["score"])

Output:

[85 90 88]

---

🟡 Accessing Individual Records

You can still use indexing for rows.

print(data[0])

Output:

('Aman', 20, 85)

---

Access Specific Field from Row

print(data[1]["score"])

Output:

90

---

🟣 Field Modification

You can also modify values using field names.

data["score"][0] = 95

print(data)

Output:

[('Aman', 20, 95)
 ('John', 22, 90)
 ('Sara', 21, 88)]

---

🔴 Filtering Using Fields

You can apply conditions directly on fields.

---

Example: Students with score > 85

result = data[data["score"] > 85]

print(result)

Output:

[('John', 22, 90)
 ('Sara', 21, 88)]

---

🟤 Multiple Field Access

You can select multiple fields together.

print(data[["name", "score"]])

Output:

[('Aman', 95)
 ('John', 90)
 ('Sara', 88)]

---

🧠 Why Use Field Access?

Field access is useful because:

• It makes data more readable
• Works like database tables
• Supports mixed data types
• Useful for structured datasets

---

⚡ Real-World Use Cases

1. Student Records

• name
• age
• marks

2. Employee Data

• id
• salary
• department

3. Sensor Data

• temperature
• humidity
• timestamp

---

📊 Field Access vs Normal Indexing

Feature	Normal Array	Structured Array
Access method	Index (0,1,2)	Field names
Data type	Single type	Multiple types
Readability	Low	High
Use case	Numerical data	Tabular data

---

🚀 Performance Note

• Structured arrays are slightly slower than regular arrays
• But they provide better readability and organization
• Ideal for small to medium structured datasets

---

⚠️ Common Mistakes

1. Using field name incorrectly

KeyError: 'score'

✔ Fix: Ensure field name matches dtype definition

---

2. Using normal arrays instead of structured arrays

✔ Field access works only with structured dtype arrays

---

🧾 Summary

NumPy field access allows you to:

• Access data using column names
• Work with structured arrays
• Filter and modify specific fields
• Handle tabular-like data efficiently

---

🏁 Conclusion

Field access in NumPy is a powerful feature for working with structured data. It brings a table-like approach to array handling, making it easier to manage real-world datasets such as employee records, student marks, and sensor data.

It is especially useful when you need clarity, structure, and named data access in numerical computing.