Products of Vectors and Matrices in NumPy

Working with vector and matrix operations is a fundamental part of scientific computing and data analysis. NumPy is a Python library that computes various types of vector and matrix products. Let's discuss how to find the inner, outer and cross products of matrices and vectors using NumPy in Python.

Inner Product

The inner product (or dot product) is obtained by multiplying corresponding elements of two arrays and summing them. For matrices, NumPy computes it row-wise.

Syntax:

numpy.inner(arr1, arr2)

Example: Below we compute the inner product of two vectors and two matrices using np.inner().

Inner product of vectors a and b =
66
Inner product of matrices x and y =
[[17 52]
 [13 62]]

Explanation:

• For vectors a and b: (2 * 3) + (6 * 10) = 66
• For matrices, each row of x is multiplied with each row of y.
• Example: First row [2,3,4] with [1,5,0] ->2 * 1 + 3 * 5 + 4 * 0 = 17

Outer Product

The outer product of two vectors creates a matrix where each element is the product of elements from both vectors. For matrices, they are flattened to 1D before applying the operation.

Syntax:

numpy.outer(arr1, arr2, out = None)

Example: Here we calculate the outer product of vectors and matrices using np.outer().

Output

Outer product of vectors a and b =
[[ 6 20]
[18 60]]

Outer product of matrices x and y =
[[ 3 45 21 9 30 24]
[ 6 90 42 18 60 48]
[ 4 60 28 12 40 32]
[ 9 135 63 27 90 72]
[ 4 60 28 12 40 32]
[ 6 90 42 18 60 48]]

Explanation:

• For vectors: Each element of a is multiplied with each element of b.
• Example: First row: [2 * 3, 2 * 10] = [6, 10]
• For matrices: Both are flattened into 1D arrays before applying pairwise multiplication.

Cross Product

The cross product is defined for 3D vectors and produces a new vector that is perpendicular to both input vectors. For 2D vectors, NumPy previously returned a scalar, but in NumPy ≥ 2.0 this behavior is deprecated. To avoid warnings and stay consistent, we explicitly represent 2D vectors as 3D by adding a zero in the z-axis (e.g., [x, y, 0]).

For matrices, the cross product is applied row by row, treating each row as a 3D vector.

Syntax:

numpy.cross(arr1 , arr2)

Example: Here we compute the cross product for vectors (converted to 3D) and for matrices row-wise.

Cross product of vectors a and b =
[ 0 0 -24]
Cross product of matrices x and y =
[[ -9 51 -32]
 [-15 3 3]]

Explanation:

• For vectors a and b: We added 0 in the z-axis, so the result is [0, 0, -24] instead of just a scalar.
• For matrices: Each row in x is crossed with the corresponding row in y.
• Example: [2,6,9] × [7,5,6] = [-39, 45, -32].