Skip to content
QuizMaker logoQuizMaker
Activity
DSA Course: Interview Patterns and Problem Solving
Module 13: Greedy Algorithms
Best Time to Buy and Sell Stock: Greedy Pattern
Maximum Subarray: Kadane Pattern
Move Zeroes: Two pointers Pattern
Contains Duplicate: Set Pattern
Valid Anagram: Frequency map Pattern
Longest Substring Without Repeating Characters: Sliding window Pattern
Valid Palindrome: Two pointers Pattern
Longest Palindromic Substring: Expand around center Pattern
Group Anagrams: Hash key Pattern
Binary Search: Classic search Pattern
Search Insert Position: Lower bound Pattern
First Bad Version: Predicate search Pattern
Search in Rotated Sorted Array: Rotated search Pattern
Find Minimum in Rotated Sorted Array: Rotated minimum Pattern
Valid Parentheses: Stack matching Pattern
Min Stack: Auxiliary stack Pattern
Daily Temperatures: Monotonic stack Pattern
Next Greater Element I: Monotonic stack Pattern
Evaluate Reverse Polish Notation: Stack evaluation Pattern
Reverse Linked List: Pointer reversal Pattern
Merge Two Sorted Lists: Dummy node Pattern
Linked List Cycle: Fast and slow pointers Pattern
Middle of the Linked List: Fast and slow pointers Pattern
Remove Nth Node From End: Two pointers Pattern
Binary Tree Traversals: DFS recursion Pattern
Maximum Depth of Binary Tree: Height recursion Pattern
Binary Tree Level Order Traversal: BFS queue Pattern
Validate Binary Search Tree: Range bounds Pattern
Lowest Common Ancestor: Recursive split Pattern
Connected Components: Adjacency DFS Pattern
Number of Islands: Grid DFS Pattern
Flood Fill: Boundary DFS Pattern
Clone Graph: Hash Map DFS Pattern
Course Schedule: Topological Sort Pattern
Union Find Components: Disjoint Set Pattern
Shortest Path in Unweighted Graph: BFS Distance Pattern
Climbing Stairs: Fibonacci DP Pattern
House Robber: Pick or Skip DP Pattern
Coin Change: Minimum Coins DP Pattern
Longest Increasing Subsequence: Binary Search DP Pattern
Longest Common Subsequence: 2D DP Pattern
0/1 Knapsack: Capacity DP Pattern
Longest Consecutive Sequence: Hash Set Pattern
Subarray Sum Equals K: Prefix Sum Hashmap Pattern
First Unique Character: Frequency Map Pattern
Find Duplicates: Frequency Map Pattern
Ransom Note: Character Availability Pattern
Sort Colors: Dutch National Flag Pattern
Next Permutation: Pivot and Suffix Reversal Pattern
Merge Intervals: Sort and Sweep Pattern
Find First and Last Position: Boundary Binary Search Pattern
Search a 2D Matrix: Flattened Binary Search Pattern
Subsets: Pick or Skip Recursion Pattern
Generate Parentheses: Valid State Backtracking Pattern
Combination Sum: Reuse Choice Backtracking Pattern
N-Queens: Constraint Backtracking Pattern
Word Search: Grid Backtracking Pattern
Kth Largest Element: Size-K Min-Heap Pattern
Top K Frequent Elements: Frequency Heap Pattern
Merge K Sorted Lists: Min-Heap Multiway Merge Pattern
Median Finder: Two Heaps Pattern
Task Scheduler: Greedy Max-Heap Pattern
Jump Game: Farthest Reach Greedy Pattern
Gas Station: Greedy Reset Pattern
Non-overlapping Intervals: Earliest End Greedy Pattern
Minimum Arrows to Burst Balloons: Interval End Greedy Pattern
Partition Labels: Last Occurrence Greedy Pattern
Single Number: XOR Cancellation Pattern
Power of Two: n and n-1 Pattern
Number of 1 Bits: Brian Kernighan Pattern
Single Number III: Rightmost Set Bit Pattern
XOR From 1 to N: Modulo Cycle Pattern
Prime Check: Square Root Trial Division Pattern
Sieve of Eratosthenes: Prime Marking Pattern
GCD: Euclidean Remainder Pattern
Binary Exponentiation: Fast Power Pattern
Modular Inverse: Extended Euclid Pattern
Implement Trie: Prefix Tree Pattern
Longest Common Prefix: Single Branch Trie Pattern
LRU Cache: Hash Map Plus Recency List Pattern
Segment Tree: Range Sum Query Pattern
Fenwick Tree: Binary Indexed Prefix Sum Pattern
CONTENTS

Partition Labels: Last Occurrence Greedy Pattern

Cut a string into maximum parts where each character appears in one part only.

DSA Course: Interview Patterns and Problem Solving
Module 13: Greedy Algorithms
dsa
greedy-algorithms
+1
May 29, 2026
23
A

Learning Outcome

After this lesson, you should be able to use last occurrence positions as dynamic boundaries for greedy cuts.

Problem Statement

Given a string, partition it into as many parts as possible so each character appears in at most one part. Return the sizes of the parts.

InputOutputWhy
s = "ababcbacadefegdehijhklij"[9,7,8]The first partition must end at index 8 because a, b, and c all appear within that boundary.

Brute Force Approach

Try every cut and validate whether characters appear across multiple parts. This is slow.

Optimized Approach

Precompute the last index of every character. While scanning, extend the current partition end to the farthest last occurrence seen so far.

Exact Pseudocode

last[ch] = final index of ch
start = 0
end = 0
for i from 0 to length(s) - 1:
  end = max(end, last[s[i]])
  if i == end:
    answer.add(end - start + 1)
    start = i + 1
return answer

Reference Code

class Solution:
    def partitionLabels(self, s):
        last = {ch: i for i, ch in enumerate(s)}
        answer = []
        start = 0
        end = 0

        for i, ch in enumerate(s):
            end = max(end, last[ch])
            if i == end:
                answer.append(end - start + 1)
                start = i + 1

        return answer

Sample Dry Run

StepStateResult
Precompute lasta ends at 8, b at 5, c at 7Boundaries are known
Scan first partitionend expands to 8Cannot cut before 8
i = 8i equals endAdd size 9
ContinueNext partitions close at 15 and 23sizes [9,7,8]

Complexity

MeasureValueReason
TimeO(n)The string is scanned twice.
SpaceO(1)For lowercase English letters, last positions use fixed space.

Edge Cases

  • A string with one repeated character becomes one partition.
  • Each character must appear in at most one output part.
  • Use the correct character range for the input.

Interview Checklist

  • Precompute last occurrence for every character.
  • Extend end while scanning the current partition.
  • Cut only when i reaches end.

FAQs

Why can we cut when i equals end?

Every character seen in the current partition has its last occurrence at or before this index.

Why not cut at first repeat?

Other characters inside the segment may appear later, so the boundary must track all last occurrences.

What is the core pattern?

Last occurrence boundary greedy.

Share this article

Share on TwitterShare on LinkedInShare on FacebookShare on WhatsAppShare on Email

Test your knowledge

Take a quick quiz based on this chapter.

mediumDSA Course
Partition Labels - Last Occurrence Greedy Pattern Practice Quiz
5 questions8 min

0 comments

Please login to comment.
No comments yet.
Lesson 5 of 5 in Module 13: Greedy Algorithms
Previous in Module 13: Greedy Algorithms
Minimum Arrows to Burst Balloons: Interval End Greedy Pattern
Next section: Module 14: Bit Manipulation
Single Number: XOR Cancellation Pattern
Module 14: Bit Manipulation
Back to DSA Course: Interview Patterns and Problem Solving
Back to moduleCategories