The problem: given an integer array nums of length n, return an array ans of length 2n where ans[i] == nums[i] and ans[i + n] == nums[i] for 0 <= i < n. In other words, nums followed by nums again.
Example: [1, 2, 1] becomes [1, 2, 1, 1, 2, 1].
Version 1: the language does it
def getConcatenation(nums):
return nums + nums
This is correct and idiomatic. In an interview it is fine to write it first, as long as you can say what it costs: a new list of size 2n is allocated and every element is copied once, so O(n) time and O(n) space.
Version 2: explicit indexing
Some interviewers want to see that you understand the index mapping, so be ready to write it by hand:
def getConcatenation(nums):
n = len(nums)
ans = [0] * (2 * n)
for i in range(n):
ans[i] = nums[i]
ans[i + n] = nums[i]
return ans
Same complexity, but now the i and i + n mapping is visible, which is the real point of the exercise.
Follow-ups worth preparing
- Repeat k times. Generalise to
ans[i + j * n] = nums[i]for j in 0..k-1, or use the modulo view:ans[t] = nums[t % n]. - Do not allocate a copy. If the caller only needs to read the result, you can expose a view that maps index t to
nums[t % n]. This is the same idea behind circular buffers and the "circular array" trick in Next Greater Element II, where you loop over 2n indices withi % ninstead of building the doubled array. - In-place in a fixed buffer. If
numsalready has capacity 2n, copy from the front into the back half; no overlap problem because the halves are disjoint.
Takeaway
Easy problems are usually a warm-up for communication: state the one-liner, state its cost, show you know the index math, and connect it to the circular-array pattern before the interviewer asks.
I work on Linkjob AI, an interview assistant that helps you practise talking through problems like this one in real time. (Disclosure: affiliated.)