This guide covers the creation, manipulation, and operations of NumPy arrays, including indexing, reshaping, concatenation, splitting, copying, and aggregation.

Creating Arrays

Using np.array()

NumPy arrays have uniform data types. If mixed types are provided, they are converted to the highest priority type: str > float > int.

import numpy as np

lst = [1, 2, 3, 4, 5, 6]
arr = np.array(lst)
print(arr, type(arr))  # [1 2 3 4 5 6] <class 'numpy.ndarray'>

# Type precedence
arr = np.array([3.14, 2, 1, 'fg'])
print(arr)  # ['3.14' '2' '1' 'fg']

Using NumPy Functions

• Ones: np.ones(shape, dtype=None)
• Zeros: np.zeros(shape, dtype=None)
• Full: np.full(shape, fill_value, dtype=None)
• Identity Matrix: np.eye(N, M=None, k=0, dtype=float)
• Linearly Spaced: np.linspace(start, stop, num=50, endpoint=True, dtype=None)
• Range: np.arange([start,]stop[, step,], dtype=None)
• Random Integers: np.random.randint(low, high=None, size=None, dtype=int)
• Standard Normal Distribution: np.random.randn(d0, d1, ..., dn)
• Normal Distribution: np.random.normal(loc=0.0, scale=1.0, size=None)
• Random Floats: np.random.random(size=None) or np.random.rand(*d0, d1, ..., dn)

n = np.ones((2, 3), dtype=int)
print(n)

n = np.full((2, 3, 4), 1)
print(n)

n = np.eye(3, 3, k=1, dtype=int)
print(n)

n = np.linspace(0, 100, num=51, dtype=int)
print(n)

n = np.arange(1, 10, 2)
print(n)  # [1 3 5 7 9]

n = np.random.randint(0, 10, (2, 3))
print(n)

n = np.random.normal(170, 5, (3, 4))
print(n)

n = np.random.rand(3, 4)
print(n)

Array Properties

n = np.array([[[1, 2], [3, 4], [5, 6]]])
print(n.ndim)  # 3
print(n.shape)  # (1, 3, 2)
print(n.size)  # 6
print(n.dtype)  # int64

Basic Operations

Indexing

n = np.array([[[1, 2, 3], [4, 5, 6]], [[7, 8, 9], [10, 11, 12]]])
print(n[0, 0, 2])  # 3
n[0, 0, 2] = 0
print(n)

n = np.arange(5)
print(n[1:3])  # [1 2]

Reshape

n = np.arange(1, 5)
n2 = np.reshape(n, (2, 2))
print(n2)

Concatenaet

n1 = np.array([[1, 2], [3, 4]])
n2 = np.array([[5, 6], [7, 8]])
print(np.concatenate((n1, n2), axis=0))

Split

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

Copy

n1 = np.arange(5)
n2 = n1.copy()
n1[0] = 100
print(n1, n2)

Transpose

n = np.array([[1, 2], [3, 4], [5, 6]])
print(n.T)

Aggregation Operations

n = np.array([1, 2, 3, 4, 5])
print(np.sum(n))  # 15
print(np.max(n))  # 5
print(np.mean(n))  # 3.0

Matrix Operations

n1 = np.array([[1, 2, 3], [2, 3, 4]])
n2 = np.array([[1, 2, 3], [3, 3, 4]])
print(n1 @ n2)

n = np.array([[1, 0, 0], [0, 1, 0], [0, 0, 1]])
print(np.linalg.inv(n))  # Inverse matrix
print(np.linalg.det(n))  # Determinant

Sorting

n1 = np.array([1, 2, 8, 6, 5])
print(np.sort(n1))  # [1 2 5 6 8]

File Operations

x = np.arange(0, 5)
y = np.arange(5, 10)

np.save('x.npy', x)
np.savez('xy.npz', xarr=x, yarr=y)

print(np.load('x.npy'))  # [0 1 2 3 4]
print(np.load('xy.npz')['yarr'])  # [5 6 7 8 9]

n = np.array([[1, 2, 3], [4, 5, 6]])
np.savetxt('n.txt', n, delimiter=',')
print(np.loadtxt('n.txt', delimiter=','))  # [[1. 2. 3.], [4. 5. 6.]]