Introduction to NumPy: Core Features, Array Creation, Operations, Indexing, and I/O

NumPy Overview

NumPy (Numerical Python) is one of the most important scientific‑computing libraries in Python, used for data analysis, scientific computing, and machine learning. Its core component is the high‑performance multidimensional array object ndarray, which supports efficient storage and operations.

Core Features of NumPy

High‑performance multidimensional array object ( ndarray)

Array operations including broadcasting and vectorized computation

Mathematical functions such as linear algebra, Fourier transforms, random number generation, etc.

Array Creation Functions

np.zeros((3, 4))      # array of zeros
np.ones((2, 2))       # array of ones
np.full((2, 3), 7)    # array filled with a constant
np.eye(4)             # identity matrix
np.arange(0, 10, 2)   # similar to range
np.linspace(0, 1, 5)   # evenly spaced numbers

Array Indexing and Slicing

a = np.array([[1, 2, 3], [4, 5, 6]])
# element access
print(a[0, 1])        # 2
# column extraction
print(a[:, 1])        # [2 5]
# row extraction
print(a[1, :])        # [4 5 6]
# boolean indexing
print(a[a > 3])        # [4 5 6]

Array Transformations and Combination

a.reshape(3, 2)      # change shape
a.T                    # transpose
np.vstack([a, a])      # vertical stack
np.hstack([a, a])      # horizontal stack
np.split(a, 2)        # split into sub‑arrays

Mathematical and Statistical Functions

np.sum(a)      # sum of all elements
np.mean(a)     # average value
np.std(a)      # standard deviation
np.min(a), np.max(a)   # min and max
np.argmin(a), np.argmax(a)   # indices of min and max

Linear Algebra Module ( numpy.linalg )

from numpy.linalg import inv, eig, solve
A = np.array([[1, 2], [3, 4]])
print(inv(A))   # matrix inverse
print(eig(A))   # eigenvalues and eigenvectors
b = np.array([5, 6])
print(solve(A, b))   # solve Ax = b

Random Number Module ( numpy.random )

np.random.rand(3, 2)          # uniform [0,1) numbers
np.random.randn(3)           # standard normal distribution
np.random.randint(1, 10, 5)   # random integers
np.random.choice([1,2,3], 4) # random selection from a list
np.random.seed(42)           # set seed for reproducibility

Input/Output Operations

np.save('my_array.npy', a)          # save as binary .npy file
np.load('my_array.npy')             # load .npy file
np.savetxt('my_array.csv', a, delimiter=',')   # save as CSV
np.loadtxt('my_array.csv', delimiter=',')      # load CSV file

Summary Mind‑Map (Logical Structure)

NumPy
├── ndarray
│   ├── attributes (shape, dtype, ndim, size)
│   └── creation (array, zeros, ones, arange, linspace)
├── operations
│   ├── arithmetic (+, -, *, /, **, exp, sqrt)
│   └── broadcasting
├── indexing & slicing
│   ├── basic indexing
│   ├── boolean indexing
│   └── fancy indexing
├── array transformation
│   ├── reshape
│   ├── transpose
│   ├── concatenate & split
├── math & statistics
│   ├── sum, mean, std, min, max
│   └── argmin, argmax
├── linear algebra
│   └── inv, eig, solve, dot
├── random module
│   └── rand, randint, choice, seed
└── I/O
    └── save, load, savetxt, loadtxt