Numerical Python (NumPy) is a powerful library for numerical computations in Python.
1. Installing and Importing NumPy
pip install numpy
import numpy as np
2. Creating Arrays
1D Array
arr = np.array([1, 2, 3, 4])
2D Array (Matrix)
arr2d = np.array([[1, 2], [3, 4]])
Zeros, Ones, and Identity Matrix
np.zeros((3, 3)) # 3x3 array of zeros np.ones((2, 4)) # 2x4 array of ones np.eye(3) # 3x3 Identity matrix
Range and Linspace
np.arange(0, 10, 2) # [0, 2, 4, 6, 8] np.linspace(0, 5, 10) # 10 points between 0 and 5
3. Array Properties
arr.shape # Shape (rows, cols) arr.size # Total number of elements arr.ndim # Number of dimensions arr.dtype # Data type of elements
4. Reshaping and Flattening Arrays
arr.reshape(2, 2) # Reshape to 2x2 arr.flatten() # Flatten to 1D array
5. Indexing and Slicing
Basic Indexing
arr[0] # First element arr[-1] # Last element arr2d[1, 1] # Element at row 1, column 1
Slicing
arr[1:3] # Elements from index 1 to 2 arr[:2] # First two elements arr[::2] # Every second element
Conditional Indexing
arr[arr > 2] # Filter elements greater than 2
6. Mathematical Operations
Element-wise Operations
arr + 2 # Add 2 to each element arr * 3 # Multiply each element by 3 np.sqrt(arr) # Square root of elements np.exp(arr) # Exponential of elements
Aggregate Functions
np.sum(arr) # Sum of all elements np.mean(arr) # Mean np.min(arr) # Minimum np.max(arr) # Maximum np.std(arr) # Standard deviation
7. Matrix Operations
np.dot(arr1, arr2) # Matrix multiplication np.transpose(arr2d) # Transpose matrix np.linalg.inv(arr2d) # Inverse of matrix
8. Random Numbers
np.random.rand(3, 3) # Uniform distribution np.random.randn(3, 3) # Standard normal distribution np.random.randint(0, 10, (2, 2)) # Random integers between 0 and 10
9. Copying and Broadcasting
Copying Arrays
b = arr.copy() # Independent copy
Broadcasting
arr + np.array([1, 2, 3, 4]) # Broadcasting addition
10. Useful NumPy Functions
| Function | Description |
|---|---|
| np.sum(arr) | Sum of elements |
| np.prod(arr) | Product of elements |
| np.cumsum(arr) | Cumulative sum |
| np.cumprod(arr) | Cumulative product |
| np.sort(arr) | Sort elements |
| np.argsort(arr) | Indices of sorted elements |
| np.unique(arr) | Unique elements |
| np.argmax(arr) | Index of max value |
| np.argmin(arr) | Index of min value |
11. Saving and Loading Data
np.save('array.npy', arr) # Save array to file
loaded_arr = np.load('array.npy') # Load array from file
12. Handling Missing Values
np.nan # Not-a-Number (NaN) np.isnan(arr) # Check for NaNs np.nanmean(arr) # Mean ignoring NaNs np.nansum(arr) # Sum ignoring NaNs
13. Boolean Operations
np.all(arr > 0) # True if all elements > 0 np.any(arr < 0) # True if any element < 0
14. Performance Tips
- Use Vectorized Operations – Faster than loops.
- Avoid Copying Arrays Unnecessarily – Use views instead.
- Preallocate Memory – Use np.empty or np.zeros to initialize large arrays.
Tips for Learning NumPy
- Practice Array Manipulation – Experiment with slicing, reshaping, and broadcasting.
- Explore NumPy Documentation – Learn more about advanced functions and optimizations.
- Apply to Real Projects – Use NumPy for data analysis, image processing, and scientific computing.
NumPy is the foundation of data science and machine learning in Python