NumPy Indexing – Access Array Elements in Python (1D, 2D, 3D Explained)

🐍 NumPy – Indexing

Indexing is one of the most important concepts in NumPy. It allows you to access individual elements from arrays quickly and efficiently.

NumPy uses zero-based indexing, meaning the first element starts at index 0.

Indexing is widely used in:

• Data analysis
• Machine learning
• Image processing
• Scientific computing
• Feature extraction

Understanding indexing is essential before learning slicing, filtering, and advanced array operations.

---

What is Indexing?

Indexing means accessing a specific element in an array using its position.

Example:

import numpy as np

arr = np.array([10, 20, 30, 40, 50])

print(arr[0])
print(arr[3])

Output:

10
40

---

🔵 Positive Indexing

Positive indexing starts from the beginning (left to right).

Index	Value
0	First element
1	Second element
2	Third element

---

Example

arr = np.array([5, 10, 15, 20])

print(arr[1])
print(arr[2])

Output:

10
15

---

🔴 Negative Indexing

Negative indexing starts from the end of the array.

Index	Value
-1	Last element
-2	Second last

---

Example

arr = np.array([10, 20, 30, 40, 50])

print(arr[-1])
print(arr[-2])

Output:

50
40

---

🟡 Indexing in 2D Arrays

In a 2D array, indexing is done using:

array[row_index, column_index]

---

Example

arr = np.array([
    [1, 2, 3],
    [4, 5, 6],
    [7, 8, 9]
])

print(arr[0, 1])
print(arr[2, 2])

Output:

2
9

---

🟢 Accessing Entire Row

print(arr[1])

Output:

[4 5 6]

✔ Returns the full row

---

🟣 Accessing Entire Column

print(arr[:, 1])

Output:

[2 5 8]

✔ : means all rows
✔ 1 means second column

---

🔵 Indexing in 3D Arrays

3D arrays represent multiple layers of data.

array[layer, row, column]

---

Example

arr = np.array([
    [[1, 2], [3, 4]],
    [[5, 6], [7, 8]]
])

print(arr[0, 1, 1])

Output:

4

---

🟠 Modifying Elements Using Indexing

Indexing is not only for reading values—you can also modify data.

arr = np.array([1, 2, 3, 4])

arr[2] = 99

print(arr)

Output:

[ 1  2 99  4]

---

🔶 Indexing Multiple Elements (Fancy Indexing)

You can access multiple elements at once.

arr = np.array([10, 20, 30, 40, 50])

print(arr[[0, 2, 4]])

Output:

[10 30 50]

---

🟤 Boolean Indexing

Boolean indexing allows filtering based on conditions.

arr = np.array([10, 15, 20, 25, 30])

print(arr[arr > 20])

Output:

[25 30]

---

⚡ Real-World Example

Extracting pixel values (Image data)

import numpy as np

image = np.array([
    [255, 128, 64],
    [100, 150, 200],
    [50,  75,  90]
])

print(image[1, 2])

Output:

200

✔ Common in image processing and computer vision

---

📊 Indexing Summary Table

Type	Description
Positive Indexing	Access from start
Negative Indexing	Access from end
2D Indexing	row, column access
3D Indexing	layer, row, column
Fancy Indexing	multiple elements
Boolean Indexing	condition-based filtering

---

🧠 Best Practices

• Always check array shape using .shape
• Use negative indexing for last elements
• Prefer boolean indexing for filtering data
• Avoid out-of-range index errors
• Combine indexing with slicing for advanced operations

---

🚫 Common Errors

Index Out of Range

IndexError: index 10 is out of bounds

✔ Fix: Ensure index is within array size

---

🏁 Summary

NumPy indexing allows you to efficiently access and modify elements in arrays. It supports:

• 1D indexing
• 2D matrix indexing
• 3D tensor indexing
• Negative indexing
• Boolean filtering
• Fancy indexing

---

🚀 Conclusion

Indexing is the foundation of working with NumPy arrays. Mastering it allows you to manipulate data efficiently and is essential for all advanced topics like slicing, filtering, and machine learning preprocessing.