Skip to content
RED SAGSkills
easy15 ptsstatisticsdevopssorting

Latency Percentiles

A dashboard reports p50, p95 and p99 response times. Use the nearest-rank method: sort the samples (milliseconds) ascending; for percentile p the rank is Math.ceil(p * n / 100) (at least 1), and the answer is the sample at that 1-based rank.

Return an array with one value per entry of percentiles, in the same order. If there are no samples, every value is null.

Example 1

Input: samples = [120,80,95,300,110,105,90,85,100,1000], percentiles = [50,90]
Output: [100,300]
Explanation: n = 10: p50 is rank 5 and p90 rank 9.

Example 2

Input: samples = [5], percentiles = [50,99,100]
Output: [5,5,5]

Constraints

  • 0 <= samples.length <= 10^5
  • 1 <= percentiles.length <= 10
  • Each percentile is an integer from 1 to 100
Hint 1

Sort numbers with (a, b) => a - b; the default sort compares strings.

Hint 2

Compute p * n before dividing by 100 so the rank is exact.

Good practice for: Site Reliability Engineer, DevOps Engineer, Cloud Engineer, Backend Developer, MLOps Engineer, Data Engineer

Run checks your code on 3 sample tests. Submit runs all 7 tests, including hidden ones, and earns points. Use console.log to debug. Python's first run can take a few seconds while it starts.