Cosine Similarity Between Vectors
Problem statement
Cosine Similarity
Implement cosine_similarity(v1, v2) that returns the cosine of the angle between two vectors, a measure of direction that ignores length.
Input: v1, v2: NumPy arrays of the same shape, non-empty and with nonzero magnitude.
Output: a float in $[-1, 1]$.
Example
Input:
v1 = np.array([1, 2, 3])
v2 = np.array([2, 4, 6])
print(round(cosine_similarity(v1, v2), 3))
Output:
1.0
Reasoning: v2 is 2 times v1, so the two vectors point in exactly the same direction.
Show hint
Which quantity measures how aligned two vectors are, before any normalization?
Relevant roles: ML Engineer, AI Engineer, LLM Engineer, Data Scientist and Applied Scientist.
Related free problems
- Multi-Head AttentionMedium
- Masked Self-AttentionMedium