Minimize L1 Distance with k Cluster Centers in Array
Overview
This question evaluates understanding of clustering and metric-based optimization, focusing on L1 distance and interval covering in a one-dimensional setting and belongs to the Machine Learning and algorithms domain with a practical application focus on algorithm design and optimization.
Atlassian
Jul 12, 2025, 6:59 PM
Data Scientist
Technical Screen
Machine Learning
13
0
1D k-Center (Minimax L1) Clustering
Problem
You are given an array of n integers on a number line and an integer k (1 ≤ k ≤ n). Place k cluster centers on the real line to minimize the maximum L1 (absolute) distance between any point and its nearest center.
- Distance between a point x and a center c is |x − c|.
- Centers may be placed anywhere on the real line (not necessarily at input points).
- Output the minimum possible maximum distance (the optimal radius).
Goal
Return the optimal radius R* such that all points can be covered by k intervals of radius R* around the chosen centers, and no smaller radius suffices.
Notes
- Think of this as covering all sorted points with k intervals of length 2R.
- Typical approach: sort, binary search on R, and use a greedy feasibility check.
- If you must return an integer and the true optimum is a half-integer, you can either output a floating value (e.g., with 1e-6 precision) or scale coordinates by 2 to keep everything integral.
Solution
Login required
Show