NumPy is one of the most powerful Python libraries used for numerical computing, data analysis, and scientific computing. One of its most essential features is NumPy array indexing, which allows you to access, modify, and manipulate elements in arrays.
Whether you’re working with 1D, 2D, or even multidimensional arrays, understanding how indexing works will significantly improve your coding efficiency and help you avoid common errors.
🔍 What is NumPy Array Indexing?
NumPy array indexing is the process of selecting elements from an array using their position (index). In NumPy, indexing works similarly to Python lists, but with more powerful features like slicing, advanced indexing, and boolean indexing.
✅ Basic Indexing in 1D Arrays
Let’s start with a simple 1D NumPy array:
pythonCopyEditimport numpy as np
arr = np.array([10, 20, 30, 40, 50])
print(arr[0]) # Output: 10
print(arr[3]) # Output: 40
- Indexing starts from 0.
- Negative indexing is also supported:
pythonCopyEditprint(arr[-1]) # Output: 50 (last element)
🔢 Indexing in 2D Arrays
A 2D array is like a matrix. You can use row and column indices to access elements.
pythonCopyEditarr2d = np.array([[1, 2, 3],
[4, 5, 6],
[7, 8, 9]])
print(arr2d[0, 1]) # Output: 2 (row 0, column 1)
print(arr2d[2, 2]) # Output: 9 (last row, last column)
You can also use negative indexing:
pythonCopyEditprint(arr2d[-1, -2]) # Output: 8
🧩 Slicing Arrays
Slicing is a powerful technique for accessing a range of elements.
1D Slicing:
pythonCopyEditarr = np.array([0, 10, 20, 30, 40, 50])
print(arr[1:4]) # Output: [10 20 30]
print(arr[:3]) # Output: [0 10 20]
print(arr[3:]) # Output: [30 40 50]
2D Slicing:
pythonCopyEditarr2d = np.array([[1, 2, 3],
[4, 5, 6],
[7, 8, 9]])
print(arr2d[:2, 1:]) # Output: [[2 3] [5 6]]
:2selects first two rows1:selects columns from index 1 to end
🔍 Boolean Indexing
Boolean indexing lets you filter elements based on conditions.
pythonCopyEditarr = np.array([5, 10, 15, 20, 25])
print(arr[arr > 15]) # Output: [20 25]
You can also combine conditions:
pythonCopyEditprint(arr[(arr > 10) & (arr < 25)]) # Output: [15 20]
🎯 Fancy Indexing (Advanced Indexing)
Fancy indexing allows you to use lists or arrays of indices to select multiple elements at once.
pythonCopyEditarr = np.array([10, 20, 30, 40, 50])
indices = [1, 3, 4]
print(arr[indices]) # Output: [20 40 50]
It also works with 2D arrays:
pythonCopyEditarr2d = np.array([[10, 11], [20, 21], [30, 31]])
print(arr2d[[0, 2], [1, 0]]) # Output: [11 30]
This selects:
- Row 0, Col 1 → 11
- Row 2, Col 0 → 30
🚫 Common Indexing Errors
- IndexError: index out of bounds
pythonCopyEditarr = np.array([1, 2, 3])
print(arr[5]) # Error ❌
- Shape mismatch in fancy indexing
Ensure index arrays match dimensions.
🧠 Tips for Using NumPy Indexing Efficiently
- Use slicing over loops for speed.
- Combine boolean and fancy indexing for powerful filters.
- Avoid unnecessary array copies—use views when possible.
- Use
.copy()if you plan to modify the selected elements independently.
❓ FAQs About NumPy Array Indexing
Q1. Can I use negative indexing in NumPy?
Yes, just like Python lists. -1 refers to the last element.
Q2. What’s the difference between slicing and fancy indexing?
Slicing returns a view; fancy indexing returns a copy.
Q3. Can I modify values using indexing?
Yes! Example:
pythonCopyEditarr[0] = 100
Q4. How do I select an entire row or column in a 2D array?
- Entire row:
arr[row_index] - Entire column:
arr[:, col_index]
Q5. Can I use multiple conditions in boolean indexing?
Yes, using & (and), | (or), ~ (not) with parentheses.
🏁 Conclusion
NumPy array indexing is a fundamental skill for any Python programmer working with data. It allows you to access and manipulate data efficiently in 1D, 2D, or multi-dimensional arrays.
From slicing to fancy and boolean indexing, NumPy provides multiple ways to access the data you need quickly. Mastering these techniques will save you time and help you write clean, efficient code—especially in data analysis, machine learning, and scientific computing.